Delete checkpoint-618

checkpoint-618/config.json

@@ -1,28 +0,0 @@
-{
-  "_name_or_path": "cyberagent/calm2-7b-chat",
-  "architectures": [
-    "LlamaForCausalLM"
-  ],
-  "attention_bias": false,
-  "bos_token_id": 0,
-  "eos_token_id": 0,
-  "hidden_act": "silu",
-  "hidden_size": 4096,
-  "initializer_range": 0.02,
-  "intermediate_size": 11008,
-  "max_position_embeddings": 32768,
-  "model_type": "llama",
-  "num_attention_heads": 32,
-  "num_hidden_layers": 32,
-  "num_key_value_heads": 32,
-  "pad_token_id": 1,
-  "pretraining_tp": 1,
-  "rms_norm_eps": 1e-06,
-  "rope_scaling": null,
-  "rope_theta": 500000,
-  "tie_word_embeddings": false,
-  "torch_dtype": "bfloat16",
-  "transformers_version": "4.34.1",
-  "use_cache": false,
-  "vocab_size": 65024
-}

checkpoint-618/generation_config.json

@@ -1,7 +0,0 @@
-{
-  "_from_model_config": true,
-  "bos_token_id": 0,
-  "eos_token_id": 0,
-  "pad_token_id": 1,
-  "transformers_version": "4.34.1"
-}

checkpoint-618/global_step618/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:cfb4514c3718fcaf0b44817b8b6282dfd700e0d12b937ecbf66871de6ff19c09
-size 28035802551

checkpoint-618/global_step618/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:6b9290cc1480f6205a0c20b10d09202764a1fa808d4fdd800776a1625bee45ce
-size 28035803191

checkpoint-618/global_step618/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f36ceff325ba6cd1535f1dfceccd0fd34a624bd63b1f33ee5278c713aee01c6f
-size 28035802743

checkpoint-618/global_step618/mp_rank_00_model_states.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ae8e77bdb20476e11de017ea64f26b80360300bcbfa293b237730d077a5329e1
-size 14017976195

checkpoint-618/latest

@@ -1 +0,0 @@
-global_step618

checkpoint-618/pytorch_model-00001-of-00002.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:245595c67e8712c740f414577e89a49e851bc1110aef9b159f85ad73d7bf63c1
-size 9976594142

checkpoint-618/pytorch_model-00002-of-00002.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9b965f08ffaf99a0f921d52572ee779b6be1f96afb3031cad865be6cbb5bfe6f
-size 4041391035

checkpoint-618/pytorch_model.bin.index.json

@@ -1,266 +0,0 @@
-{
-  "metadata": {
-    "total_size": 14017896448
-  },
-  "weight_map": {
-    "lm_head.weight": "pytorch_model-00002-of-00002.bin",
-    "model.embed_tokens.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.0.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.0.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.1.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.1.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.10.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.10.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.10.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.11.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.11.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.11.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.12.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.12.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.12.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.13.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.13.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.13.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.14.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.14.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.14.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.15.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.15.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.15.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.16.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.16.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.16.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.17.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.17.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.17.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.18.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.18.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.18.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.19.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.19.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.19.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.2.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.2.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.20.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.20.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.20.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.21.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.21.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.21.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.22.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.22.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.22.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.23.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.23.mlp.swiglu.w12.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.23.mlp.swiglu.w3.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.24.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.24.mlp.swiglu.w12.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.24.mlp.swiglu.w3.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.24.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.25.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.25.mlp.swiglu.w12.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.25.mlp.swiglu.w3.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.25.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.26.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.26.mlp.swiglu.w12.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.26.mlp.swiglu.w3.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.26.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.26.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.26.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.26.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.27.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.27.mlp.swiglu.w12.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.27.mlp.swiglu.w3.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.27.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.27.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.27.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.27.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.28.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.28.mlp.swiglu.w12.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.28.mlp.swiglu.w3.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.28.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.28.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.28.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.28.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.28.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.29.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.29.mlp.swiglu.w12.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.29.mlp.swiglu.w3.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.29.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.29.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.29.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.29.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.29.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.3.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.3.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.30.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.30.mlp.swiglu.w12.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.30.mlp.swiglu.w3.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.30.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.30.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.30.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.30.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.30.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.31.input_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.31.mlp.swiglu.w12.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.31.mlp.swiglu.w3.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.31.post_attention_layernorm.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.31.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.31.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.31.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.31.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
-    "model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.4.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.4.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.5.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.5.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.5.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.6.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.6.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.6.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.7.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.7.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.7.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.8.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.8.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.8.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.9.input_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.9.mlp.swiglu.w12.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.9.mlp.swiglu.w3.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
-    "model.norm.weight": "pytorch_model-00002-of-00002.bin"
-  }
-}

checkpoint-618/rng_state_0.pth

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4f4f4eabd3d3209be5ecfa7748b59c9bcebe66f8280e04423295c3adb56fdda8
-size 16631

checkpoint-618/rng_state_1.pth

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0e5d459c15b3659d339b29f90d9c6d4fdbf6c828b592cb47110d9ed8c71e113f
-size 16631

checkpoint-618/rng_state_2.pth

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:7b4bd6bae9c008f0e9d18f4569f4dddb1adc43f27c07518ce3f88803299dc53b
-size 16631

checkpoint-618/trainer_state.json

@@ -1,3887 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 2.9611650485436893,
-  "eval_steps": 31,
-  "global_step": 618,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.0,
-      "loss": 4.59,
-      "step": 1
-    },
-    {
-      "epoch": 0.0,
-      "eval_loss": 4.240383148193359,
-      "eval_runtime": 7.311,
-      "eval_samples_per_second": 164.273,
-      "eval_steps_per_second": 54.849,
-      "step": 1
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 2.0000000000000003e-06,
-      "loss": 4.252,
-      "step": 2
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 4.2054,
-      "step": 3
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 6e-06,
-      "loss": 4.1624,
-      "step": 4
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 8.000000000000001e-06,
-      "loss": 3.9787,
-      "step": 5
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1e-05,
-      "loss": 3.7979,
-      "step": 6
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 1.2e-05,
-      "loss": 3.8982,
-      "step": 7
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.4000000000000001e-05,
-      "loss": 3.805,
-      "step": 8
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 1.6000000000000003e-05,
-      "loss": 3.7176,
-      "step": 9
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 1.8e-05,
-      "loss": 3.4755,
-      "step": 10
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 2e-05,
-      "loss": 3.6401,
-      "step": 11
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.2000000000000003e-05,
-      "loss": 3.5615,
-      "step": 12
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 2.4e-05,
-      "loss": 3.5286,
-      "step": 13
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.6000000000000002e-05,
-      "loss": 3.5437,
-      "step": 14
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 2.8000000000000003e-05,
-      "loss": 3.5163,
-      "step": 15
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 3e-05,
-      "loss": 3.4108,
-      "step": 16
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 3.2000000000000005e-05,
-      "loss": 3.3637,
-      "step": 17
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 3.4000000000000007e-05,
-      "loss": 3.3538,
-      "step": 18
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 3.6e-05,
-      "loss": 3.3819,
-      "step": 19
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 3.8e-05,
-      "loss": 3.2511,
-      "step": 20
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 4e-05,
-      "loss": 3.3211,
-      "step": 21
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 4.2e-05,
-      "loss": 3.2764,
-      "step": 22
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 4.4000000000000006e-05,
-      "loss": 3.0653,
-      "step": 23
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 4.600000000000001e-05,
-      "loss": 3.0859,
-      "step": 24
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 4.8e-05,
-      "loss": 3.0804,
-      "step": 25
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5e-05,
-      "loss": 2.9774,
-      "step": 26
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 5.2000000000000004e-05,
-      "loss": 2.9269,
-      "step": 27
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.4000000000000005e-05,
-      "loss": 3.0926,
-      "step": 28
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 5.6000000000000006e-05,
-      "loss": 2.9725,
-      "step": 29
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 5.8e-05,
-      "loss": 3.0293,
-      "step": 30
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 6e-05,
-      "loss": 3.0903,
-      "step": 31
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 2.9962990283966064,
-      "eval_runtime": 7.3671,
-      "eval_samples_per_second": 163.022,
-      "eval_steps_per_second": 54.431,
-      "step": 31
-    },
-    {
-      "epoch": 0.16,
-      "learning_rate": 6.2e-05,
-      "loss": 2.9903,
-      "step": 32
-    },
-    {
-      "epoch": 0.16,
-      "learning_rate": 6.400000000000001e-05,
-      "loss": 3.0196,
-      "step": 33
-    },
-    {
-      "epoch": 0.17,
-      "learning_rate": 6.6e-05,
-      "loss": 3.0288,
-      "step": 34
-    },
-    {
-      "epoch": 0.17,
-      "learning_rate": 6.800000000000001e-05,
-      "loss": 3.0071,
-      "step": 35
-    },
-    {
-      "epoch": 0.17,
-      "learning_rate": 7e-05,
-      "loss": 3.0393,
-      "step": 36
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 7.2e-05,
-      "loss": 2.9937,
-      "step": 37
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 7.4e-05,
-      "loss": 2.9988,
-      "step": 38
-    },
-    {
-      "epoch": 0.19,
-      "learning_rate": 7.6e-05,
-      "loss": 2.9331,
-      "step": 39
-    },
-    {
-      "epoch": 0.19,
-      "learning_rate": 7.800000000000001e-05,
-      "loss": 3.0414,
-      "step": 40
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 8e-05,
-      "loss": 3.0237,
-      "step": 41
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 8.2e-05,
-      "loss": 2.9664,
-      "step": 42
-    },
-    {
-      "epoch": 0.21,
-      "learning_rate": 8.4e-05,
-      "loss": 2.8639,
-      "step": 43
-    },
-    {
-      "epoch": 0.21,
-      "learning_rate": 8.6e-05,
-      "loss": 2.8562,
-      "step": 44
-    },
-    {
-      "epoch": 0.22,
-      "learning_rate": 8.800000000000001e-05,
-      "loss": 2.9632,
-      "step": 45
-    },
-    {
-      "epoch": 0.22,
-      "learning_rate": 9e-05,
-      "loss": 2.946,
-      "step": 46
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 9.200000000000001e-05,
-      "loss": 2.8428,
-      "step": 47
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 9.4e-05,
-      "loss": 2.9827,
-      "step": 48
-    },
-    {
-      "epoch": 0.24,
-      "learning_rate": 9.6e-05,
-      "loss": 2.9512,
-      "step": 49
-    },
-    {
-      "epoch": 0.24,
-      "learning_rate": 9.8e-05,
-      "loss": 2.8997,
-      "step": 50
-    },
-    {
-      "epoch": 0.25,
-      "learning_rate": 0.0001,
-      "loss": 2.9762,
-      "step": 51
-    },
-    {
-      "epoch": 0.25,
-      "learning_rate": 0.00010200000000000001,
-      "loss": 3.0429,
-      "step": 52
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 0.00010400000000000001,
-      "loss": 3.0223,
-      "step": 53
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 0.00010600000000000002,
-      "loss": 3.0007,
-      "step": 54
-    },
-    {
-      "epoch": 0.27,
-      "learning_rate": 0.00010800000000000001,
-      "loss": 3.0436,
-      "step": 55
-    },
-    {
-      "epoch": 0.27,
-      "learning_rate": 0.00011000000000000002,
-      "loss": 3.0151,
-      "step": 56
-    },
-    {
-      "epoch": 0.28,
-      "learning_rate": 0.00011200000000000001,
-      "loss": 2.9909,
-      "step": 57
-    },
-    {
-      "epoch": 0.28,
-      "learning_rate": 0.00011399999999999999,
-      "loss": 2.9942,
-      "step": 58
-    },
-    {
-      "epoch": 0.29,
-      "learning_rate": 0.000116,
-      "loss": 3.0098,
-      "step": 59
-    },
-    {
-      "epoch": 0.29,
-      "learning_rate": 0.000118,
-      "loss": 3.0353,
-      "step": 60
-    },
-    {
-      "epoch": 0.3,
-      "learning_rate": 0.00012,
-      "loss": 3.0671,
-      "step": 61
-    },
-    {
-      "epoch": 0.3,
-      "learning_rate": 0.000122,
-      "loss": 2.9824,
-      "step": 62
-    },
-    {
-      "epoch": 0.3,
-      "eval_loss": 3.022158145904541,
-      "eval_runtime": 7.3702,
-      "eval_samples_per_second": 162.953,
-      "eval_steps_per_second": 54.408,
-      "step": 62
-    },
-    {
-      "epoch": 0.31,
-      "learning_rate": 0.000124,
-      "loss": 3.0207,
-      "step": 63
-    },
-    {
-      "epoch": 0.31,
-      "learning_rate": 0.000126,
-      "loss": 2.9048,
-      "step": 64
-    },
-    {
-      "epoch": 0.32,
-      "learning_rate": 0.00012800000000000002,
-      "loss": 3.0518,
-      "step": 65
-    },
-    {
-      "epoch": 0.32,
-      "learning_rate": 0.00013000000000000002,
-      "loss": 3.0854,
-      "step": 66
-    },
-    {
-      "epoch": 0.33,
-      "learning_rate": 0.000132,
-      "loss": 3.0317,
-      "step": 67
-    },
-    {
-      "epoch": 0.33,
-      "learning_rate": 0.000134,
-      "loss": 3.0313,
-      "step": 68
-    },
-    {
-      "epoch": 0.33,
-      "learning_rate": 0.00013600000000000003,
-      "loss": 3.0753,
-      "step": 69
-    },
-    {
-      "epoch": 0.34,
-      "learning_rate": 0.000138,
-      "loss": 2.9999,
-      "step": 70
-    },
-    {
-      "epoch": 0.34,
-      "learning_rate": 0.00014,
-      "loss": 3.0423,
-      "step": 71
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 0.000142,
-      "loss": 2.9642,
-      "step": 72
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 0.000144,
-      "loss": 2.9575,
-      "step": 73
-    },
-    {
-      "epoch": 0.36,
-      "learning_rate": 0.000146,
-      "loss": 2.9854,
-      "step": 74
-    },
-    {
-      "epoch": 0.36,
-      "learning_rate": 0.000148,
-      "loss": 2.9729,
-      "step": 75
-    },
-    {
-      "epoch": 0.37,
-      "learning_rate": 0.00015000000000000001,
-      "loss": 2.9176,
-      "step": 76
-    },
-    {
-      "epoch": 0.37,
-      "learning_rate": 0.000152,
-      "loss": 2.947,
-      "step": 77
-    },
-    {
-      "epoch": 0.38,
-      "learning_rate": 0.000154,
-      "loss": 3.0542,
-      "step": 78
-    },
-    {
-      "epoch": 0.38,
-      "learning_rate": 0.00015600000000000002,
-      "loss": 3.0718,
-      "step": 79
-    },
-    {
-      "epoch": 0.39,
-      "learning_rate": 0.00015800000000000002,
-      "loss": 3.027,
-      "step": 80
-    },
-    {
-      "epoch": 0.39,
-      "learning_rate": 0.00016,
-      "loss": 3.1764,
-      "step": 81
-    },
-    {
-      "epoch": 0.4,
-      "learning_rate": 0.000162,
-      "loss": 3.1091,
-      "step": 82
-    },
-    {
-      "epoch": 0.4,
-      "learning_rate": 0.000164,
-      "loss": 3.0931,
-      "step": 83
-    },
-    {
-      "epoch": 0.41,
-      "learning_rate": 0.000166,
-      "loss": 3.2712,
-      "step": 84
-    },
-    {
-      "epoch": 0.41,
-      "learning_rate": 0.000168,
-      "loss": 3.3353,
-      "step": 85
-    },
-    {
-      "epoch": 0.42,
-      "learning_rate": 0.00017,
-      "loss": 3.4876,
-      "step": 86
-    },
-    {
-      "epoch": 0.42,
-      "learning_rate": 0.000172,
-      "loss": 3.3383,
-      "step": 87
-    },
-    {
-      "epoch": 0.43,
-      "learning_rate": 0.000174,
-      "loss": 3.1497,
-      "step": 88
-    },
-    {
-      "epoch": 0.43,
-      "learning_rate": 0.00017600000000000002,
-      "loss": 3.1029,
-      "step": 89
-    },
-    {
-      "epoch": 0.44,
-      "learning_rate": 0.00017800000000000002,
-      "loss": 3.1484,
-      "step": 90
-    },
-    {
-      "epoch": 0.44,
-      "learning_rate": 0.00018,
-      "loss": 3.1156,
-      "step": 91
-    },
-    {
-      "epoch": 0.45,
-      "learning_rate": 0.000182,
-      "loss": 3.2557,
-      "step": 92
-    },
-    {
-      "epoch": 0.45,
-      "learning_rate": 0.00018400000000000003,
-      "loss": 3.173,
-      "step": 93
-    },
-    {
-      "epoch": 0.45,
-      "eval_loss": 3.150902032852173,
-      "eval_runtime": 7.5676,
-      "eval_samples_per_second": 158.703,
-      "eval_steps_per_second": 52.989,
-      "step": 93
-    },
-    {
-      "epoch": 0.46,
-      "learning_rate": 0.00018600000000000002,
-      "loss": 3.128,
-      "step": 94
-    },
-    {
-      "epoch": 0.46,
-      "learning_rate": 0.000188,
-      "loss": 3.146,
-      "step": 95
-    },
-    {
-      "epoch": 0.47,
-      "learning_rate": 0.00019,
-      "loss": 3.194,
-      "step": 96
-    },
-    {
-      "epoch": 0.47,
-      "learning_rate": 0.000192,
-      "loss": 3.0987,
-      "step": 97
-    },
-    {
-      "epoch": 0.48,
-      "learning_rate": 0.000194,
-      "loss": 3.2405,
-      "step": 98
-    },
-    {
-      "epoch": 0.48,
-      "learning_rate": 0.000196,
-      "loss": 3.1568,
-      "step": 99
-    },
-    {
-      "epoch": 0.49,
-      "learning_rate": 0.00019800000000000002,
-      "loss": 3.1488,
-      "step": 100
-    },
-    {
-      "epoch": 0.49,
-      "learning_rate": 0.0002,
-      "loss": 3.2105,
-      "step": 101
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 0.0001996138996138996,
-      "loss": 3.2575,
-      "step": 102
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 0.00019922779922779924,
-      "loss": 3.1921,
-      "step": 103
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 0.00019884169884169884,
-      "loss": 3.2369,
-      "step": 104
-    },
-    {
-      "epoch": 0.51,
-      "learning_rate": 0.00019845559845559847,
-      "loss": 3.1031,
-      "step": 105
-    },
-    {
-      "epoch": 0.51,
-      "learning_rate": 0.00019806949806949807,
-      "loss": 3.2618,
-      "step": 106
-    },
-    {
-      "epoch": 0.52,
-      "learning_rate": 0.0001976833976833977,
-      "loss": 3.2034,
-      "step": 107
-    },
-    {
-      "epoch": 0.52,
-      "learning_rate": 0.0001972972972972973,
-      "loss": 3.2094,
-      "step": 108
-    },
-    {
-      "epoch": 0.53,
-      "learning_rate": 0.00019691119691119693,
-      "loss": 3.1894,
-      "step": 109
-    },
-    {
-      "epoch": 0.53,
-      "learning_rate": 0.00019652509652509653,
-      "loss": 3.1614,
-      "step": 110
-    },
-    {
-      "epoch": 0.54,
-      "learning_rate": 0.00019613899613899616,
-      "loss": 3.176,
-      "step": 111
-    },
-    {
-      "epoch": 0.54,
-      "learning_rate": 0.00019575289575289576,
-      "loss": 3.2153,
-      "step": 112
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 0.0001953667953667954,
-      "loss": 3.0923,
-      "step": 113
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 0.000194980694980695,
-      "loss": 3.2878,
-      "step": 114
-    },
-    {
-      "epoch": 0.56,
-      "learning_rate": 0.00019459459459459462,
-      "loss": 3.0605,
-      "step": 115
-    },
-    {
-      "epoch": 0.56,
-      "learning_rate": 0.00019420849420849422,
-      "loss": 3.1282,
-      "step": 116
-    },
-    {
-      "epoch": 0.57,
-      "learning_rate": 0.00019382239382239382,
-      "loss": 3.1204,
-      "step": 117
-    },
-    {
-      "epoch": 0.57,
-      "learning_rate": 0.00019343629343629345,
-      "loss": 3.0932,
-      "step": 118
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 0.00019305019305019305,
-      "loss": 3.2913,
-      "step": 119
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 0.00019266409266409268,
-      "loss": 3.1809,
-      "step": 120
-    },
-    {
-      "epoch": 0.59,
-      "learning_rate": 0.00019227799227799228,
-      "loss": 3.2408,
-      "step": 121
-    },
-    {
-      "epoch": 0.59,
-      "learning_rate": 0.0001918918918918919,
-      "loss": 3.8133,
-      "step": 122
-    },
-    {
-      "epoch": 0.6,
-      "learning_rate": 0.0001915057915057915,
-      "loss": 3.1869,
-      "step": 123
-    },
-    {
-      "epoch": 0.6,
-      "learning_rate": 0.00019111969111969114,
-      "loss": 3.1053,
-      "step": 124
-    },
-    {
-      "epoch": 0.6,
-      "eval_loss": 3.205463409423828,
-      "eval_runtime": 7.3677,
-      "eval_samples_per_second": 163.009,
-      "eval_steps_per_second": 54.427,
-      "step": 124
-    },
-    {
-      "epoch": 0.61,
-      "learning_rate": 0.00019073359073359074,
-      "loss": 3.1206,
-      "step": 125
-    },
-    {
-      "epoch": 0.61,
-      "learning_rate": 0.00019034749034749037,
-      "loss": 3.1233,
-      "step": 126
-    },
-    {
-      "epoch": 0.62,
-      "learning_rate": 0.00018996138996138997,
-      "loss": 3.0673,
-      "step": 127
-    },
-    {
-      "epoch": 0.62,
-      "learning_rate": 0.0001895752895752896,
-      "loss": 3.1314,
-      "step": 128
-    },
-    {
-      "epoch": 0.63,
-      "learning_rate": 0.0001891891891891892,
-      "loss": 3.1997,
-      "step": 129
-    },
-    {
-      "epoch": 0.63,
-      "learning_rate": 0.0001888030888030888,
-      "loss": 3.1298,
-      "step": 130
-    },
-    {
-      "epoch": 0.64,
-      "learning_rate": 0.00018841698841698843,
-      "loss": 3.1821,
-      "step": 131
-    },
-    {
-      "epoch": 0.64,
-      "learning_rate": 0.00018803088803088803,
-      "loss": 3.2418,
-      "step": 132
-    },
-    {
-      "epoch": 0.65,
-      "learning_rate": 0.00018764478764478766,
-      "loss": 3.1543,
-      "step": 133
-    },
-    {
-      "epoch": 0.65,
-      "learning_rate": 0.00018725868725868726,
-      "loss": 3.2136,
-      "step": 134
-    },
-    {
-      "epoch": 0.66,
-      "learning_rate": 0.0001868725868725869,
-      "loss": 3.3314,
-      "step": 135
-    },
-    {
-      "epoch": 0.66,
-      "learning_rate": 0.0001864864864864865,
-      "loss": 3.2328,
-      "step": 136
-    },
-    {
-      "epoch": 0.67,
-      "learning_rate": 0.00018610038610038612,
-      "loss": 3.2225,
-      "step": 137
-    },
-    {
-      "epoch": 0.67,
-      "learning_rate": 0.00018571428571428572,
-      "loss": 3.1159,
-      "step": 138
-    },
-    {
-      "epoch": 0.67,
-      "learning_rate": 0.00018532818532818535,
-      "loss": 3.0339,
-      "step": 139
-    },
-    {
-      "epoch": 0.68,
-      "learning_rate": 0.00018494208494208495,
-      "loss": 3.2672,
-      "step": 140
-    },
-    {
-      "epoch": 0.68,
-      "learning_rate": 0.00018455598455598458,
-      "loss": 3.1237,
-      "step": 141
-    },
-    {
-      "epoch": 0.69,
-      "learning_rate": 0.00018416988416988418,
-      "loss": 3.1692,
-      "step": 142
-    },
-    {
-      "epoch": 0.69,
-      "learning_rate": 0.0001837837837837838,
-      "loss": 3.3243,
-      "step": 143
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 0.0001833976833976834,
-      "loss": 3.0264,
-      "step": 144
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 0.000183011583011583,
-      "loss": 3.2045,
-      "step": 145
-    },
-    {
-      "epoch": 0.71,
-      "learning_rate": 0.00018262548262548264,
-      "loss": 3.1966,
-      "step": 146
-    },
-    {
-      "epoch": 0.71,
-      "learning_rate": 0.00018223938223938224,
-      "loss": 3.0587,
-      "step": 147
-    },
-    {
-      "epoch": 0.72,
-      "learning_rate": 0.00018185328185328187,
-      "loss": 3.2979,
-      "step": 148
-    },
-    {
-      "epoch": 0.72,
-      "learning_rate": 0.00018146718146718147,
-      "loss": 3.1549,
-      "step": 149
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 0.0001810810810810811,
-      "loss": 3.1682,
-      "step": 150
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 0.0001806949806949807,
-      "loss": 3.3214,
-      "step": 151
-    },
-    {
-      "epoch": 0.74,
-      "learning_rate": 0.00018030888030888032,
-      "loss": 3.1783,
-      "step": 152
-    },
-    {
-      "epoch": 0.74,
-      "learning_rate": 0.00017992277992277993,
-      "loss": 3.2268,
-      "step": 153
-    },
-    {
-      "epoch": 0.75,
-      "learning_rate": 0.00017953667953667955,
-      "loss": 3.2843,
-      "step": 154
-    },
-    {
-      "epoch": 0.75,
-      "learning_rate": 0.00017915057915057916,
-      "loss": 3.3124,
-      "step": 155
-    },
-    {
-      "epoch": 0.75,
-      "eval_loss": 3.1898298263549805,
-      "eval_runtime": 7.5884,
-      "eval_samples_per_second": 158.268,
-      "eval_steps_per_second": 52.844,
-      "step": 155
-    },
-    {
-      "epoch": 0.76,
-      "learning_rate": 0.00017876447876447878,
-      "loss": 3.1855,
-      "step": 156
-    },
-    {
-      "epoch": 0.76,
-      "learning_rate": 0.00017837837837837839,
-      "loss": 3.2096,
-      "step": 157
-    },
-    {
-      "epoch": 0.77,
-      "learning_rate": 0.00017799227799227801,
-      "loss": 3.2174,
-      "step": 158
-    },
-    {
-      "epoch": 0.77,
-      "learning_rate": 0.00017760617760617762,
-      "loss": 3.2775,
-      "step": 159
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 0.00017722007722007722,
-      "loss": 3.2065,
-      "step": 160
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 0.00017683397683397684,
-      "loss": 3.2905,
-      "step": 161
-    },
-    {
-      "epoch": 0.79,
-      "learning_rate": 0.00017644787644787645,
-      "loss": 3.1591,
-      "step": 162
-    },
-    {
-      "epoch": 0.79,
-      "learning_rate": 0.00017606177606177607,
-      "loss": 3.2721,
-      "step": 163
-    },
-    {
-      "epoch": 0.8,
-      "learning_rate": 0.00017567567567567568,
-      "loss": 3.1743,
-      "step": 164
-    },
-    {
-      "epoch": 0.8,
-      "learning_rate": 0.0001752895752895753,
-      "loss": 3.234,
-      "step": 165
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 0.0001749034749034749,
-      "loss": 3.2775,
-      "step": 166
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 0.00017451737451737453,
-      "loss": 3.2317,
-      "step": 167
-    },
-    {
-      "epoch": 0.82,
-      "learning_rate": 0.00017413127413127413,
-      "loss": 3.0691,
-      "step": 168
-    },
-    {
-      "epoch": 0.82,
-      "learning_rate": 0.00017374517374517376,
-      "loss": 3.1793,
-      "step": 169
-    },
-    {
-      "epoch": 0.83,
-      "learning_rate": 0.00017335907335907336,
-      "loss": 3.2259,
-      "step": 170
-    },
-    {
-      "epoch": 0.83,
-      "learning_rate": 0.000172972972972973,
-      "loss": 3.1813,
-      "step": 171
-    },
-    {
-      "epoch": 0.83,
-      "learning_rate": 0.0001725868725868726,
-      "loss": 3.2416,
-      "step": 172
-    },
-    {
-      "epoch": 0.84,
-      "learning_rate": 0.0001722007722007722,
-      "loss": 3.2016,
-      "step": 173
-    },
-    {
-      "epoch": 0.84,
-      "learning_rate": 0.00017181467181467182,
-      "loss": 3.1766,
-      "step": 174
-    },
-    {
-      "epoch": 0.85,
-      "learning_rate": 0.00017142857142857143,
-      "loss": 3.0861,
-      "step": 175
-    },
-    {
-      "epoch": 0.85,
-      "learning_rate": 0.00017104247104247105,
-      "loss": 3.2104,
-      "step": 176
-    },
-    {
-      "epoch": 0.86,
-      "learning_rate": 0.00017065637065637065,
-      "loss": 3.273,
-      "step": 177
-    },
-    {
-      "epoch": 0.86,
-      "learning_rate": 0.00017027027027027028,
-      "loss": 3.2371,
-      "step": 178
-    },
-    {
-      "epoch": 0.87,
-      "learning_rate": 0.00016988416988416988,
-      "loss": 3.2654,
-      "step": 179
-    },
-    {
-      "epoch": 0.87,
-      "learning_rate": 0.0001694980694980695,
-      "loss": 3.1812,
-      "step": 180
-    },
-    {
-      "epoch": 0.88,
-      "learning_rate": 0.00016911196911196911,
-      "loss": 3.2781,
-      "step": 181
-    },
-    {
-      "epoch": 0.88,
-      "learning_rate": 0.00016872586872586874,
-      "loss": 3.1611,
-      "step": 182
-    },
-    {
-      "epoch": 0.89,
-      "learning_rate": 0.00016833976833976834,
-      "loss": 3.0902,
-      "step": 183
-    },
-    {
-      "epoch": 0.89,
-      "learning_rate": 0.00016795366795366797,
-      "loss": 3.2414,
-      "step": 184
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 0.00016756756756756757,
-      "loss": 3.1472,
-      "step": 185
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 0.0001671814671814672,
-      "loss": 3.1761,
-      "step": 186
-    },
-    {
-      "epoch": 0.9,
-      "eval_loss": 3.1637256145477295,
-      "eval_runtime": 7.3669,
-      "eval_samples_per_second": 163.026,
-      "eval_steps_per_second": 54.432,
-      "step": 186
-    },
-    {
-      "epoch": 0.91,
-      "learning_rate": 0.0001667953667953668,
-      "loss": 3.1409,
-      "step": 187
-    },
-    {
-      "epoch": 0.91,
-      "learning_rate": 0.0001664092664092664,
-      "loss": 3.2262,
-      "step": 188
-    },
-    {
-      "epoch": 0.92,
-      "learning_rate": 0.00016602316602316603,
-      "loss": 3.105,
-      "step": 189
-    },
-    {
-      "epoch": 0.92,
-      "learning_rate": 0.00016563706563706563,
-      "loss": 3.2596,
-      "step": 190
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 0.00016525096525096526,
-      "loss": 3.1528,
-      "step": 191
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 0.00016486486486486486,
-      "loss": 3.1561,
-      "step": 192
-    },
-    {
-      "epoch": 0.94,
-      "learning_rate": 0.0001644787644787645,
-      "loss": 3.2552,
-      "step": 193
-    },
-    {
-      "epoch": 0.94,
-      "learning_rate": 0.0001640926640926641,
-      "loss": 3.0347,
-      "step": 194
-    },
-    {
-      "epoch": 0.95,
-      "learning_rate": 0.00016370656370656372,
-      "loss": 3.0418,
-      "step": 195
-    },
-    {
-      "epoch": 0.95,
-      "learning_rate": 0.00016332046332046332,
-      "loss": 3.0838,
-      "step": 196
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 0.00016293436293436295,
-      "loss": 3.1867,
-      "step": 197
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 0.00016254826254826255,
-      "loss": 3.0373,
-      "step": 198
-    },
-    {
-      "epoch": 0.97,
-      "learning_rate": 0.00016216216216216218,
-      "loss": 2.9896,
-      "step": 199
-    },
-    {
-      "epoch": 0.97,
-      "learning_rate": 0.00016177606177606178,
-      "loss": 3.1511,
-      "step": 200
-    },
-    {
-      "epoch": 0.98,
-      "learning_rate": 0.0001613899613899614,
-      "loss": 3.1029,
-      "step": 201
-    },
-    {
-      "epoch": 0.98,
-      "learning_rate": 0.000161003861003861,
-      "loss": 3.2193,
-      "step": 202
-    },
-    {
-      "epoch": 0.99,
-      "learning_rate": 0.0001606177606177606,
-      "loss": 3.2214,
-      "step": 203
-    },
-    {
-      "epoch": 0.99,
-      "learning_rate": 0.00016023166023166024,
-      "loss": 3.1428,
-      "step": 204
-    },
-    {
-      "epoch": 1.0,
-      "learning_rate": 0.00015984555984555984,
-      "loss": 3.1259,
-      "step": 205
-    },
-    {
-      "epoch": 1.0,
-      "learning_rate": 0.00015945945945945947,
-      "loss": 3.2007,
-      "step": 206
-    },
-    {
-      "epoch": 1.0,
-      "learning_rate": 0.00015907335907335907,
-      "loss": 3.1123,
-      "step": 207
-    },
-    {
-      "epoch": 1.01,
-      "learning_rate": 0.0001586872586872587,
-      "loss": 3.3417,
-      "step": 208
-    },
-    {
-      "epoch": 1.01,
-      "learning_rate": 0.0001583011583011583,
-      "loss": 3.089,
-      "step": 209
-    },
-    {
-      "epoch": 1.02,
-      "learning_rate": 0.00015791505791505793,
-      "loss": 3.0972,
-      "step": 210
-    },
-    {
-      "epoch": 1.0,
-      "learning_rate": 0.00015752895752895753,
-      "loss": 2.1341,
-      "step": 211
-    },
-    {
-      "epoch": 1.01,
-      "learning_rate": 0.00015714285714285716,
-      "loss": 1.9415,
-      "step": 212
-    },
-    {
-      "epoch": 1.01,
-      "learning_rate": 0.00015675675675675676,
-      "loss": 1.959,
-      "step": 213
-    },
-    {
-      "epoch": 1.02,
-      "learning_rate": 0.0001563706563706564,
-      "loss": 1.857,
-      "step": 214
-    },
-    {
-      "epoch": 1.02,
-      "learning_rate": 0.000155984555984556,
-      "loss": 1.8255,
-      "step": 215
-    },
-    {
-      "epoch": 1.03,
-      "learning_rate": 0.00015559845559845562,
-      "loss": 1.6538,
-      "step": 216
-    },
-    {
-      "epoch": 1.03,
-      "learning_rate": 0.00015521235521235522,
-      "loss": 1.9162,
-      "step": 217
-    },
-    {
-      "epoch": 1.03,
-      "eval_loss": 3.399775981903076,
-      "eval_runtime": 7.3711,
-      "eval_samples_per_second": 162.933,
-      "eval_steps_per_second": 54.401,
-      "step": 217
-    },
-    {
-      "epoch": 1.04,
-      "learning_rate": 0.00015482625482625482,
-      "loss": 1.8293,
-      "step": 218
-    },
-    {
-      "epoch": 1.04,
-      "learning_rate": 0.00015444015444015445,
-      "loss": 1.8539,
-      "step": 219
-    },
-    {
-      "epoch": 1.05,
-      "learning_rate": 0.00015405405405405405,
-      "loss": 1.7888,
-      "step": 220
-    },
-    {
-      "epoch": 1.05,
-      "learning_rate": 0.00015366795366795368,
-      "loss": 1.7813,
-      "step": 221
-    },
-    {
-      "epoch": 1.06,
-      "learning_rate": 0.00015328185328185328,
-      "loss": 1.8911,
-      "step": 222
-    },
-    {
-      "epoch": 1.06,
-      "learning_rate": 0.0001528957528957529,
-      "loss": 1.839,
-      "step": 223
-    },
-    {
-      "epoch": 1.07,
-      "learning_rate": 0.0001525096525096525,
-      "loss": 1.8223,
-      "step": 224
-    },
-    {
-      "epoch": 1.07,
-      "learning_rate": 0.00015212355212355214,
-      "loss": 1.825,
-      "step": 225
-    },
-    {
-      "epoch": 1.08,
-      "learning_rate": 0.00015173745173745174,
-      "loss": 1.7962,
-      "step": 226
-    },
-    {
-      "epoch": 1.08,
-      "learning_rate": 0.00015135135135135137,
-      "loss": 1.8117,
-      "step": 227
-    },
-    {
-      "epoch": 1.09,
-      "learning_rate": 0.00015096525096525097,
-      "loss": 1.9061,
-      "step": 228
-    },
-    {
-      "epoch": 1.09,
-      "learning_rate": 0.0001505791505791506,
-      "loss": 1.8982,
-      "step": 229
-    },
-    {
-      "epoch": 1.1,
-      "learning_rate": 0.0001501930501930502,
-      "loss": 1.9087,
-      "step": 230
-    },
-    {
-      "epoch": 1.1,
-      "learning_rate": 0.0001498069498069498,
-      "loss": 1.9664,
-      "step": 231
-    },
-    {
-      "epoch": 1.11,
-      "learning_rate": 0.00014942084942084943,
-      "loss": 1.7921,
-      "step": 232
-    },
-    {
-      "epoch": 1.11,
-      "learning_rate": 0.00014903474903474903,
-      "loss": 1.9163,
-      "step": 233
-    },
-    {
-      "epoch": 1.12,
-      "learning_rate": 0.00014864864864864866,
-      "loss": 1.8759,
-      "step": 234
-    },
-    {
-      "epoch": 1.12,
-      "learning_rate": 0.00014826254826254826,
-      "loss": 1.9262,
-      "step": 235
-    },
-    {
-      "epoch": 1.13,
-      "learning_rate": 0.0001478764478764479,
-      "loss": 1.9063,
-      "step": 236
-    },
-    {
-      "epoch": 1.13,
-      "learning_rate": 0.0001474903474903475,
-      "loss": 1.8577,
-      "step": 237
-    },
-    {
-      "epoch": 1.14,
-      "learning_rate": 0.00014710424710424712,
-      "loss": 1.7452,
-      "step": 238
-    },
-    {
-      "epoch": 1.14,
-      "learning_rate": 0.00014671814671814672,
-      "loss": 1.9344,
-      "step": 239
-    },
-    {
-      "epoch": 1.15,
-      "learning_rate": 0.00014633204633204635,
-      "loss": 1.7575,
-      "step": 240
-    },
-    {
-      "epoch": 1.15,
-      "learning_rate": 0.00014594594594594595,
-      "loss": 1.7707,
-      "step": 241
-    },
-    {
-      "epoch": 1.16,
-      "learning_rate": 0.00014555984555984558,
-      "loss": 1.8945,
-      "step": 242
-    },
-    {
-      "epoch": 1.16,
-      "learning_rate": 0.00014517374517374518,
-      "loss": 1.8379,
-      "step": 243
-    },
-    {
-      "epoch": 1.17,
-      "learning_rate": 0.0001447876447876448,
-      "loss": 1.9021,
-      "step": 244
-    },
-    {
-      "epoch": 1.17,
-      "learning_rate": 0.0001444015444015444,
-      "loss": 1.844,
-      "step": 245
-    },
-    {
-      "epoch": 1.17,
-      "learning_rate": 0.000144015444015444,
-      "loss": 1.9396,
-      "step": 246
-    },
-    {
-      "epoch": 1.18,
-      "learning_rate": 0.00014362934362934364,
-      "loss": 2.0305,
-      "step": 247
-    },
-    {
-      "epoch": 1.18,
-      "learning_rate": 0.00014324324324324324,
-      "loss": 1.8985,
-      "step": 248
-    },
-    {
-      "epoch": 1.18,
-      "eval_loss": 3.330674409866333,
-      "eval_runtime": 7.364,
-      "eval_samples_per_second": 163.091,
-      "eval_steps_per_second": 54.454,
-      "step": 248
-    },
-    {
-      "epoch": 1.19,
-      "learning_rate": 0.00014285714285714287,
-      "loss": 1.8457,
-      "step": 249
-    },
-    {
-      "epoch": 1.19,
-      "learning_rate": 0.00014247104247104247,
-      "loss": 1.8213,
-      "step": 250
-    },
-    {
-      "epoch": 1.2,
-      "learning_rate": 0.0001420849420849421,
-      "loss": 1.7586,
-      "step": 251
-    },
-    {
-      "epoch": 1.2,
-      "learning_rate": 0.0001416988416988417,
-      "loss": 1.8669,
-      "step": 252
-    },
-    {
-      "epoch": 1.21,
-      "learning_rate": 0.00014131274131274133,
-      "loss": 1.9476,
-      "step": 253
-    },
-    {
-      "epoch": 1.21,
-      "learning_rate": 0.00014092664092664093,
-      "loss": 1.8525,
-      "step": 254
-    },
-    {
-      "epoch": 1.22,
-      "learning_rate": 0.00014054054054054056,
-      "loss": 2.0163,
-      "step": 255
-    },
-    {
-      "epoch": 1.22,
-      "learning_rate": 0.00014015444015444016,
-      "loss": 1.9186,
-      "step": 256
-    },
-    {
-      "epoch": 1.23,
-      "learning_rate": 0.00013976833976833979,
-      "loss": 1.9528,
-      "step": 257
-    },
-    {
-      "epoch": 1.23,
-      "learning_rate": 0.0001393822393822394,
-      "loss": 2.2483,
-      "step": 258
-    },
-    {
-      "epoch": 1.24,
-      "learning_rate": 0.00013899613899613902,
-      "loss": 1.8889,
-      "step": 259
-    },
-    {
-      "epoch": 1.24,
-      "learning_rate": 0.00013861003861003862,
-      "loss": 2.0137,
-      "step": 260
-    },
-    {
-      "epoch": 1.25,
-      "learning_rate": 0.00013822393822393822,
-      "loss": 1.9397,
-      "step": 261
-    },
-    {
-      "epoch": 1.25,
-      "learning_rate": 0.00013783783783783785,
-      "loss": 1.8241,
-      "step": 262
-    },
-    {
-      "epoch": 1.26,
-      "learning_rate": 0.00013745173745173745,
-      "loss": 1.9685,
-      "step": 263
-    },
-    {
-      "epoch": 1.26,
-      "learning_rate": 0.00013706563706563708,
-      "loss": 1.9909,
-      "step": 264
-    },
-    {
-      "epoch": 1.27,
-      "learning_rate": 0.00013667953667953668,
-      "loss": 1.8653,
-      "step": 265
-    },
-    {
-      "epoch": 1.27,
-      "learning_rate": 0.0001362934362934363,
-      "loss": 1.9163,
-      "step": 266
-    },
-    {
-      "epoch": 1.28,
-      "learning_rate": 0.0001359073359073359,
-      "loss": 1.9765,
-      "step": 267
-    },
-    {
-      "epoch": 1.28,
-      "learning_rate": 0.00013552123552123554,
-      "loss": 1.788,
-      "step": 268
-    },
-    {
-      "epoch": 1.29,
-      "learning_rate": 0.00013513513513513514,
-      "loss": 1.8103,
-      "step": 269
-    },
-    {
-      "epoch": 1.29,
-      "learning_rate": 0.00013474903474903477,
-      "loss": 2.0086,
-      "step": 270
-    },
-    {
-      "epoch": 1.3,
-      "learning_rate": 0.00013436293436293437,
-      "loss": 1.9448,
-      "step": 271
-    },
-    {
-      "epoch": 1.3,
-      "learning_rate": 0.000133976833976834,
-      "loss": 1.8598,
-      "step": 272
-    },
-    {
-      "epoch": 1.31,
-      "learning_rate": 0.0001335907335907336,
-      "loss": 2.0792,
-      "step": 273
-    },
-    {
-      "epoch": 1.31,
-      "learning_rate": 0.0001332046332046332,
-      "loss": 1.7766,
-      "step": 274
-    },
-    {
-      "epoch": 1.32,
-      "learning_rate": 0.00013281853281853283,
-      "loss": 1.9329,
-      "step": 275
-    },
-    {
-      "epoch": 1.32,
-      "learning_rate": 0.00013243243243243243,
-      "loss": 1.9933,
-      "step": 276
-    },
-    {
-      "epoch": 1.33,
-      "learning_rate": 0.00013204633204633206,
-      "loss": 1.9371,
-      "step": 277
-    },
-    {
-      "epoch": 1.33,
-      "learning_rate": 0.00013166023166023166,
-      "loss": 1.9629,
-      "step": 278
-    },
-    {
-      "epoch": 1.33,
-      "learning_rate": 0.00013127413127413129,
-      "loss": 2.0488,
-      "step": 279
-    },
-    {
-      "epoch": 1.33,
-      "eval_loss": 3.333745241165161,
-      "eval_runtime": 7.361,
-      "eval_samples_per_second": 163.157,
-      "eval_steps_per_second": 54.476,
-      "step": 279
-    },
-    {
-      "epoch": 1.34,
-      "learning_rate": 0.0001308880308880309,
-      "loss": 2.0148,
-      "step": 280
-    },
-    {
-      "epoch": 1.34,
-      "learning_rate": 0.00013050193050193052,
-      "loss": 1.8416,
-      "step": 281
-    },
-    {
-      "epoch": 1.35,
-      "learning_rate": 0.00013011583011583012,
-      "loss": 2.1004,
-      "step": 282
-    },
-    {
-      "epoch": 1.35,
-      "learning_rate": 0.00012972972972972974,
-      "loss": 1.8308,
-      "step": 283
-    },
-    {
-      "epoch": 1.36,
-      "learning_rate": 0.00012934362934362935,
-      "loss": 1.9441,
-      "step": 284
-    },
-    {
-      "epoch": 1.36,
-      "learning_rate": 0.00012895752895752897,
-      "loss": 2.083,
-      "step": 285
-    },
-    {
-      "epoch": 1.37,
-      "learning_rate": 0.00012857142857142858,
-      "loss": 1.8198,
-      "step": 286
-    },
-    {
-      "epoch": 1.37,
-      "learning_rate": 0.0001281853281853282,
-      "loss": 2.0069,
-      "step": 287
-    },
-    {
-      "epoch": 1.38,
-      "learning_rate": 0.0001277992277992278,
-      "loss": 2.0146,
-      "step": 288
-    },
-    {
-      "epoch": 1.38,
-      "learning_rate": 0.0001274131274131274,
-      "loss": 1.8554,
-      "step": 289
-    },
-    {
-      "epoch": 1.39,
-      "learning_rate": 0.00012702702702702703,
-      "loss": 1.972,
-      "step": 290
-    },
-    {
-      "epoch": 1.39,
-      "learning_rate": 0.00012664092664092664,
-      "loss": 1.9583,
-      "step": 291
-    },
-    {
-      "epoch": 1.4,
-      "learning_rate": 0.00012625482625482626,
-      "loss": 1.8567,
-      "step": 292
-    },
-    {
-      "epoch": 1.4,
-      "learning_rate": 0.00012586872586872587,
-      "loss": 2.0031,
-      "step": 293
-    },
-    {
-      "epoch": 1.41,
-      "learning_rate": 0.0001254826254826255,
-      "loss": 1.9725,
-      "step": 294
-    },
-    {
-      "epoch": 1.41,
-      "learning_rate": 0.0001250965250965251,
-      "loss": 1.9517,
-      "step": 295
-    },
-    {
-      "epoch": 1.42,
-      "learning_rate": 0.00012471042471042472,
-      "loss": 1.7436,
-      "step": 296
-    },
-    {
-      "epoch": 1.42,
-      "learning_rate": 0.00012432432432432433,
-      "loss": 1.9968,
-      "step": 297
-    },
-    {
-      "epoch": 1.43,
-      "learning_rate": 0.00012393822393822395,
-      "loss": 1.8299,
-      "step": 298
-    },
-    {
-      "epoch": 1.43,
-      "learning_rate": 0.00012355212355212355,
-      "loss": 2.1024,
-      "step": 299
-    },
-    {
-      "epoch": 1.44,
-      "learning_rate": 0.00012316602316602318,
-      "loss": 1.8099,
-      "step": 300
-    },
-    {
-      "epoch": 1.44,
-      "learning_rate": 0.00012277992277992278,
-      "loss": 1.9761,
-      "step": 301
-    },
-    {
-      "epoch": 1.45,
-      "learning_rate": 0.0001223938223938224,
-      "loss": 2.1201,
-      "step": 302
-    },
-    {
-      "epoch": 1.45,
-      "learning_rate": 0.00012200772200772201,
-      "loss": 1.9268,
-      "step": 303
-    },
-    {
-      "epoch": 1.46,
-      "learning_rate": 0.00012162162162162163,
-      "loss": 1.8136,
-      "step": 304
-    },
-    {
-      "epoch": 1.46,
-      "learning_rate": 0.00012123552123552124,
-      "loss": 2.0362,
-      "step": 305
-    },
-    {
-      "epoch": 1.47,
-      "learning_rate": 0.00012084942084942086,
-      "loss": 2.0653,
-      "step": 306
-    },
-    {
-      "epoch": 1.47,
-      "learning_rate": 0.00012046332046332047,
-      "loss": 2.022,
-      "step": 307
-    },
-    {
-      "epoch": 1.48,
-      "learning_rate": 0.00012007722007722009,
-      "loss": 1.9317,
-      "step": 308
-    },
-    {
-      "epoch": 1.48,
-      "learning_rate": 0.0001196911196911197,
-      "loss": 2.0455,
-      "step": 309
-    },
-    {
-      "epoch": 1.49,
-      "learning_rate": 0.00011930501930501932,
-      "loss": 2.0812,
-      "step": 310
-    },
-    {
-      "epoch": 1.49,
-      "eval_loss": 3.341298818588257,
-      "eval_runtime": 7.3644,
-      "eval_samples_per_second": 163.081,
-      "eval_steps_per_second": 54.451,
-      "step": 310
-    },
-    {
-      "epoch": 1.49,
-      "learning_rate": 0.00011891891891891893,
-      "loss": 2.0609,
-      "step": 311
-    },
-    {
-      "epoch": 1.5,
-      "learning_rate": 0.00011853281853281855,
-      "loss": 1.9708,
-      "step": 312
-    },
-    {
-      "epoch": 1.5,
-      "learning_rate": 0.00011814671814671816,
-      "loss": 1.9968,
-      "step": 313
-    },
-    {
-      "epoch": 1.5,
-      "learning_rate": 0.00011776061776061778,
-      "loss": 2.0283,
-      "step": 314
-    },
-    {
-      "epoch": 1.51,
-      "learning_rate": 0.00011737451737451739,
-      "loss": 2.0142,
-      "step": 315
-    },
-    {
-      "epoch": 1.51,
-      "learning_rate": 0.00011698841698841701,
-      "loss": 2.026,
-      "step": 316
-    },
-    {
-      "epoch": 1.52,
-      "learning_rate": 0.0001166023166023166,
-      "loss": 2.1965,
-      "step": 317
-    },
-    {
-      "epoch": 1.52,
-      "learning_rate": 0.00011621621621621621,
-      "loss": 1.984,
-      "step": 318
-    },
-    {
-      "epoch": 1.53,
-      "learning_rate": 0.00011583011583011582,
-      "loss": 2.0699,
-      "step": 319
-    },
-    {
-      "epoch": 1.53,
-      "learning_rate": 0.00011544401544401544,
-      "loss": 1.864,
-      "step": 320
-    },
-    {
-      "epoch": 1.54,
-      "learning_rate": 0.00011505791505791505,
-      "loss": 2.0219,
-      "step": 321
-    },
-    {
-      "epoch": 1.54,
-      "learning_rate": 0.00011467181467181467,
-      "loss": 1.9162,
-      "step": 322
-    },
-    {
-      "epoch": 1.55,
-      "learning_rate": 0.00011428571428571428,
-      "loss": 1.9092,
-      "step": 323
-    },
-    {
-      "epoch": 1.55,
-      "learning_rate": 0.0001138996138996139,
-      "loss": 2.0932,
-      "step": 324
-    },
-    {
-      "epoch": 1.56,
-      "learning_rate": 0.00011351351351351351,
-      "loss": 2.0975,
-      "step": 325
-    },
-    {
-      "epoch": 1.56,
-      "learning_rate": 0.00011312741312741313,
-      "loss": 2.1674,
-      "step": 326
-    },
-    {
-      "epoch": 1.57,
-      "learning_rate": 0.00011274131274131274,
-      "loss": 1.8444,
-      "step": 327
-    },
-    {
-      "epoch": 1.57,
-      "learning_rate": 0.00011235521235521236,
-      "loss": 1.9696,
-      "step": 328
-    },
-    {
-      "epoch": 1.58,
-      "learning_rate": 0.00011196911196911197,
-      "loss": 1.943,
-      "step": 329
-    },
-    {
-      "epoch": 1.58,
-      "learning_rate": 0.00011158301158301159,
-      "loss": 2.1044,
-      "step": 330
-    },
-    {
-      "epoch": 1.59,
-      "learning_rate": 0.0001111969111969112,
-      "loss": 2.2068,
-      "step": 331
-    },
-    {
-      "epoch": 1.59,
-      "learning_rate": 0.00011081081081081082,
-      "loss": 2.0958,
-      "step": 332
-    },
-    {
-      "epoch": 1.6,
-      "learning_rate": 0.00011042471042471043,
-      "loss": 1.9789,
-      "step": 333
-    },
-    {
-      "epoch": 1.6,
-      "learning_rate": 0.00011003861003861005,
-      "loss": 1.8663,
-      "step": 334
-    },
-    {
-      "epoch": 1.61,
-      "learning_rate": 0.00010965250965250966,
-      "loss": 2.0499,
-      "step": 335
-    },
-    {
-      "epoch": 1.61,
-      "learning_rate": 0.00010926640926640928,
-      "loss": 1.935,
-      "step": 336
-    },
-    {
-      "epoch": 1.62,
-      "learning_rate": 0.00010888030888030889,
-      "loss": 2.0021,
-      "step": 337
-    },
-    {
-      "epoch": 1.62,
-      "learning_rate": 0.0001084942084942085,
-      "loss": 1.953,
-      "step": 338
-    },
-    {
-      "epoch": 1.63,
-      "learning_rate": 0.00010810810810810812,
-      "loss": 2.0466,
-      "step": 339
-    },
-    {
-      "epoch": 1.63,
-      "learning_rate": 0.00010772200772200774,
-      "loss": 1.9709,
-      "step": 340
-    },
-    {
-      "epoch": 1.64,
-      "learning_rate": 0.00010733590733590735,
-      "loss": 1.8884,
-      "step": 341
-    },
-    {
-      "epoch": 1.64,
-      "eval_loss": 3.323758602142334,
-      "eval_runtime": 7.3665,
-      "eval_samples_per_second": 163.036,
-      "eval_steps_per_second": 54.436,
-      "step": 341
-    },
-    {
-      "epoch": 1.64,
-      "learning_rate": 0.00010694980694980697,
-      "loss": 2.0557,
-      "step": 342
-    },
-    {
-      "epoch": 1.65,
-      "learning_rate": 0.00010656370656370658,
-      "loss": 2.0345,
-      "step": 343
-    },
-    {
-      "epoch": 1.65,
-      "learning_rate": 0.0001061776061776062,
-      "loss": 1.8173,
-      "step": 344
-    },
-    {
-      "epoch": 1.66,
-      "learning_rate": 0.00010579150579150581,
-      "loss": 1.9598,
-      "step": 345
-    },
-    {
-      "epoch": 1.66,
-      "learning_rate": 0.0001054054054054054,
-      "loss": 2.0323,
-      "step": 346
-    },
-    {
-      "epoch": 1.67,
-      "learning_rate": 0.00010501930501930501,
-      "loss": 1.9284,
-      "step": 347
-    },
-    {
-      "epoch": 1.67,
-      "learning_rate": 0.00010463320463320463,
-      "loss": 2.1235,
-      "step": 348
-    },
-    {
-      "epoch": 1.67,
-      "learning_rate": 0.00010424710424710424,
-      "loss": 1.9426,
-      "step": 349
-    },
-    {
-      "epoch": 1.68,
-      "learning_rate": 0.00010386100386100386,
-      "loss": 1.8692,
-      "step": 350
-    },
-    {
-      "epoch": 1.68,
-      "learning_rate": 0.00010347490347490347,
-      "loss": 1.9559,
-      "step": 351
-    },
-    {
-      "epoch": 1.69,
-      "learning_rate": 0.00010308880308880309,
-      "loss": 2.0407,
-      "step": 352
-    },
-    {
-      "epoch": 1.69,
-      "learning_rate": 0.0001027027027027027,
-      "loss": 1.9356,
-      "step": 353
-    },
-    {
-      "epoch": 1.7,
-      "learning_rate": 0.00010231660231660232,
-      "loss": 1.9055,
-      "step": 354
-    },
-    {
-      "epoch": 1.7,
-      "learning_rate": 0.00010193050193050193,
-      "loss": 1.9776,
-      "step": 355
-    },
-    {
-      "epoch": 1.71,
-      "learning_rate": 0.00010154440154440155,
-      "loss": 1.8996,
-      "step": 356
-    },
-    {
-      "epoch": 1.71,
-      "learning_rate": 0.00010115830115830116,
-      "loss": 1.8893,
-      "step": 357
-    },
-    {
-      "epoch": 1.72,
-      "learning_rate": 0.00010077220077220078,
-      "loss": 1.9621,
-      "step": 358
-    },
-    {
-      "epoch": 1.72,
-      "learning_rate": 0.00010038610038610039,
-      "loss": 1.9447,
-      "step": 359
-    },
-    {
-      "epoch": 1.73,
-      "learning_rate": 0.0001,
-      "loss": 1.9646,
-      "step": 360
-    },
-    {
-      "epoch": 1.73,
-      "learning_rate": 9.961389961389962e-05,
-      "loss": 1.9459,
-      "step": 361
-    },
-    {
-      "epoch": 1.74,
-      "learning_rate": 9.922779922779923e-05,
-      "loss": 2.0168,
-      "step": 362
-    },
-    {
-      "epoch": 1.74,
-      "learning_rate": 9.884169884169885e-05,
-      "loss": 2.0021,
-      "step": 363
-    },
-    {
-      "epoch": 1.75,
-      "learning_rate": 9.845559845559846e-05,
-      "loss": 1.7642,
-      "step": 364
-    },
-    {
-      "epoch": 1.75,
-      "learning_rate": 9.806949806949808e-05,
-      "loss": 2.0138,
-      "step": 365
-    },
-    {
-      "epoch": 1.76,
-      "learning_rate": 9.76833976833977e-05,
-      "loss": 2.0108,
-      "step": 366
-    },
-    {
-      "epoch": 1.76,
-      "learning_rate": 9.729729729729731e-05,
-      "loss": 2.0955,
-      "step": 367
-    },
-    {
-      "epoch": 1.77,
-      "learning_rate": 9.691119691119691e-05,
-      "loss": 2.1705,
-      "step": 368
-    },
-    {
-      "epoch": 1.77,
-      "learning_rate": 9.652509652509652e-05,
-      "loss": 1.8494,
-      "step": 369
-    },
-    {
-      "epoch": 1.78,
-      "learning_rate": 9.613899613899614e-05,
-      "loss": 1.8676,
-      "step": 370
-    },
-    {
-      "epoch": 1.78,
-      "learning_rate": 9.575289575289575e-05,
-      "loss": 1.821,
-      "step": 371
-    },
-    {
-      "epoch": 1.79,
-      "learning_rate": 9.536679536679537e-05,
-      "loss": 2.0281,
-      "step": 372
-    },
-    {
-      "epoch": 1.79,
-      "eval_loss": 3.2718801498413086,
-      "eval_runtime": 7.3652,
-      "eval_samples_per_second": 163.063,
-      "eval_steps_per_second": 54.445,
-      "step": 372
-    },
-    {
-      "epoch": 1.79,
-      "learning_rate": 9.498069498069498e-05,
-      "loss": 2.1556,
-      "step": 373
-    },
-    {
-      "epoch": 1.8,
-      "learning_rate": 9.45945945945946e-05,
-      "loss": 1.9643,
-      "step": 374
-    },
-    {
-      "epoch": 1.8,
-      "learning_rate": 9.420849420849421e-05,
-      "loss": 2.0287,
-      "step": 375
-    },
-    {
-      "epoch": 1.81,
-      "learning_rate": 9.382239382239383e-05,
-      "loss": 1.927,
-      "step": 376
-    },
-    {
-      "epoch": 1.81,
-      "learning_rate": 9.343629343629344e-05,
-      "loss": 1.9838,
-      "step": 377
-    },
-    {
-      "epoch": 1.82,
-      "learning_rate": 9.305019305019306e-05,
-      "loss": 1.9065,
-      "step": 378
-    },
-    {
-      "epoch": 1.82,
-      "learning_rate": 9.266409266409267e-05,
-      "loss": 2.056,
-      "step": 379
-    },
-    {
-      "epoch": 1.83,
-      "learning_rate": 9.227799227799229e-05,
-      "loss": 1.9277,
-      "step": 380
-    },
-    {
-      "epoch": 1.83,
-      "learning_rate": 9.18918918918919e-05,
-      "loss": 1.7797,
-      "step": 381
-    },
-    {
-      "epoch": 1.83,
-      "learning_rate": 9.15057915057915e-05,
-      "loss": 1.8818,
-      "step": 382
-    },
-    {
-      "epoch": 1.84,
-      "learning_rate": 9.111969111969112e-05,
-      "loss": 1.947,
-      "step": 383
-    },
-    {
-      "epoch": 1.84,
-      "learning_rate": 9.073359073359073e-05,
-      "loss": 1.942,
-      "step": 384
-    },
-    {
-      "epoch": 1.85,
-      "learning_rate": 9.034749034749035e-05,
-      "loss": 1.9998,
-      "step": 385
-    },
-    {
-      "epoch": 1.85,
-      "learning_rate": 8.996138996138996e-05,
-      "loss": 1.8805,
-      "step": 386
-    },
-    {
-      "epoch": 1.86,
-      "learning_rate": 8.957528957528958e-05,
-      "loss": 1.8903,
-      "step": 387
-    },
-    {
-      "epoch": 1.86,
-      "learning_rate": 8.918918918918919e-05,
-      "loss": 1.9189,
-      "step": 388
-    },
-    {
-      "epoch": 1.87,
-      "learning_rate": 8.880308880308881e-05,
-      "loss": 2.0308,
-      "step": 389
-    },
-    {
-      "epoch": 1.87,
-      "learning_rate": 8.841698841698842e-05,
-      "loss": 2.0768,
-      "step": 390
-    },
-    {
-      "epoch": 1.88,
-      "learning_rate": 8.803088803088804e-05,
-      "loss": 1.9168,
-      "step": 391
-    },
-    {
-      "epoch": 1.88,
-      "learning_rate": 8.764478764478765e-05,
-      "loss": 1.8967,
-      "step": 392
-    },
-    {
-      "epoch": 1.89,
-      "learning_rate": 8.725868725868727e-05,
-      "loss": 1.9347,
-      "step": 393
-    },
-    {
-      "epoch": 1.89,
-      "learning_rate": 8.687258687258688e-05,
-      "loss": 1.8273,
-      "step": 394
-    },
-    {
-      "epoch": 1.9,
-      "learning_rate": 8.64864864864865e-05,
-      "loss": 1.9801,
-      "step": 395
-    },
-    {
-      "epoch": 1.9,
-      "learning_rate": 8.61003861003861e-05,
-      "loss": 2.0002,
-      "step": 396
-    },
-    {
-      "epoch": 1.91,
-      "learning_rate": 8.571428571428571e-05,
-      "loss": 2.0318,
-      "step": 397
-    },
-    {
-      "epoch": 1.91,
-      "learning_rate": 8.532818532818533e-05,
-      "loss": 1.8399,
-      "step": 398
-    },
-    {
-      "epoch": 1.92,
-      "learning_rate": 8.494208494208494e-05,
-      "loss": 1.8956,
-      "step": 399
-    },
-    {
-      "epoch": 1.92,
-      "learning_rate": 8.455598455598456e-05,
-      "loss": 2.0156,
-      "step": 400
-    },
-    {
-      "epoch": 1.93,
-      "learning_rate": 8.416988416988417e-05,
-      "loss": 1.9499,
-      "step": 401
-    },
-    {
-      "epoch": 1.93,
-      "learning_rate": 8.378378378378379e-05,
-      "loss": 1.8823,
-      "step": 402
-    },
-    {
-      "epoch": 1.94,
-      "learning_rate": 8.33976833976834e-05,
-      "loss": 2.1344,
-      "step": 403
-    },
-    {
-      "epoch": 1.94,
-      "eval_loss": 3.2487306594848633,
-      "eval_runtime": 7.3645,
-      "eval_samples_per_second": 163.08,
-      "eval_steps_per_second": 54.451,
-      "step": 403
-    },
-    {
-      "epoch": 1.94,
-      "learning_rate": 8.301158301158302e-05,
-      "loss": 1.9887,
-      "step": 404
-    },
-    {
-      "epoch": 1.95,
-      "learning_rate": 8.262548262548263e-05,
-      "loss": 2.0445,
-      "step": 405
-    },
-    {
-      "epoch": 1.95,
-      "learning_rate": 8.223938223938225e-05,
-      "loss": 1.8847,
-      "step": 406
-    },
-    {
-      "epoch": 1.96,
-      "learning_rate": 8.185328185328186e-05,
-      "loss": 1.8461,
-      "step": 407
-    },
-    {
-      "epoch": 1.96,
-      "learning_rate": 8.146718146718148e-05,
-      "loss": 1.9106,
-      "step": 408
-    },
-    {
-      "epoch": 1.97,
-      "learning_rate": 8.108108108108109e-05,
-      "loss": 2.0067,
-      "step": 409
-    },
-    {
-      "epoch": 1.97,
-      "learning_rate": 8.06949806949807e-05,
-      "loss": 1.9705,
-      "step": 410
-    },
-    {
-      "epoch": 1.98,
-      "learning_rate": 8.03088803088803e-05,
-      "loss": 1.8092,
-      "step": 411
-    },
-    {
-      "epoch": 1.98,
-      "learning_rate": 7.992277992277992e-05,
-      "loss": 1.8563,
-      "step": 412
-    },
-    {
-      "epoch": 1.99,
-      "learning_rate": 7.953667953667954e-05,
-      "loss": 1.8833,
-      "step": 413
-    },
-    {
-      "epoch": 1.99,
-      "learning_rate": 7.915057915057915e-05,
-      "loss": 1.9905,
-      "step": 414
-    },
-    {
-      "epoch": 2.0,
-      "learning_rate": 7.876447876447877e-05,
-      "loss": 2.0448,
-      "step": 415
-    },
-    {
-      "epoch": 2.0,
-      "learning_rate": 7.837837837837838e-05,
-      "loss": 1.9066,
-      "step": 416
-    },
-    {
-      "epoch": 2.0,
-      "learning_rate": 7.7992277992278e-05,
-      "loss": 1.8585,
-      "step": 417
-    },
-    {
-      "epoch": 2.01,
-      "learning_rate": 7.760617760617761e-05,
-      "loss": 2.0163,
-      "step": 418
-    },
-    {
-      "epoch": 2.01,
-      "learning_rate": 7.722007722007723e-05,
-      "loss": 1.8571,
-      "step": 419
-    },
-    {
-      "epoch": 2.02,
-      "learning_rate": 7.683397683397684e-05,
-      "loss": 2.0083,
-      "step": 420
-    },
-    {
-      "epoch": 2.0,
-      "learning_rate": 7.644787644787645e-05,
-      "loss": 0.6158,
-      "step": 421
-    },
-    {
-      "epoch": 2.01,
-      "learning_rate": 7.606177606177607e-05,
-      "loss": 0.7386,
-      "step": 422
-    },
-    {
-      "epoch": 2.01,
-      "learning_rate": 7.567567567567568e-05,
-      "loss": 0.7067,
-      "step": 423
-    },
-    {
-      "epoch": 2.02,
-      "learning_rate": 7.52895752895753e-05,
-      "loss": 0.6173,
-      "step": 424
-    },
-    {
-      "epoch": 2.02,
-      "learning_rate": 7.49034749034749e-05,
-      "loss": 0.5876,
-      "step": 425
-    },
-    {
-      "epoch": 2.03,
-      "learning_rate": 7.451737451737452e-05,
-      "loss": 0.5948,
-      "step": 426
-    },
-    {
-      "epoch": 2.03,
-      "learning_rate": 7.413127413127413e-05,
-      "loss": 0.5593,
-      "step": 427
-    },
-    {
-      "epoch": 2.04,
-      "learning_rate": 7.374517374517374e-05,
-      "loss": 0.5989,
-      "step": 428
-    },
-    {
-      "epoch": 2.04,
-      "learning_rate": 7.335907335907336e-05,
-      "loss": 0.5699,
-      "step": 429
-    },
-    {
-      "epoch": 2.05,
-      "learning_rate": 7.297297297297297e-05,
-      "loss": 0.5719,
-      "step": 430
-    },
-    {
-      "epoch": 2.05,
-      "learning_rate": 7.258687258687259e-05,
-      "loss": 0.4928,
-      "step": 431
-    },
-    {
-      "epoch": 2.06,
-      "learning_rate": 7.22007722007722e-05,
-      "loss": 0.4713,
-      "step": 432
-    },
-    {
-      "epoch": 2.06,
-      "learning_rate": 7.181467181467182e-05,
-      "loss": 0.6161,
-      "step": 433
-    },
-    {
-      "epoch": 2.07,
-      "learning_rate": 7.142857142857143e-05,
-      "loss": 0.566,
-      "step": 434
-    },
-    {
-      "epoch": 2.07,
-      "eval_loss": 4.280820369720459,
-      "eval_runtime": 7.3682,
-      "eval_samples_per_second": 162.998,
-      "eval_steps_per_second": 54.423,
-      "step": 434
-    },
-    {
-      "epoch": 2.07,
-      "learning_rate": 7.104247104247105e-05,
-      "loss": 0.5182,
-      "step": 435
-    },
-    {
-      "epoch": 2.08,
-      "learning_rate": 7.065637065637066e-05,
-      "loss": 0.6347,
-      "step": 436
-    },
-    {
-      "epoch": 2.08,
-      "learning_rate": 7.027027027027028e-05,
-      "loss": 0.6002,
-      "step": 437
-    },
-    {
-      "epoch": 2.09,
-      "learning_rate": 6.988416988416989e-05,
-      "loss": 0.5696,
-      "step": 438
-    },
-    {
-      "epoch": 2.09,
-      "learning_rate": 6.949806949806951e-05,
-      "loss": 0.5535,
-      "step": 439
-    },
-    {
-      "epoch": 2.1,
-      "learning_rate": 6.911196911196911e-05,
-      "loss": 0.5263,
-      "step": 440
-    },
-    {
-      "epoch": 2.1,
-      "learning_rate": 6.872586872586872e-05,
-      "loss": 0.5342,
-      "step": 441
-    },
-    {
-      "epoch": 2.11,
-      "learning_rate": 6.833976833976834e-05,
-      "loss": 0.4946,
-      "step": 442
-    },
-    {
-      "epoch": 2.11,
-      "learning_rate": 6.795366795366795e-05,
-      "loss": 0.5402,
-      "step": 443
-    },
-    {
-      "epoch": 2.12,
-      "learning_rate": 6.756756756756757e-05,
-      "loss": 0.5005,
-      "step": 444
-    },
-    {
-      "epoch": 2.12,
-      "learning_rate": 6.718146718146718e-05,
-      "loss": 0.6038,
-      "step": 445
-    },
-    {
-      "epoch": 2.13,
-      "learning_rate": 6.67953667953668e-05,
-      "loss": 0.5123,
-      "step": 446
-    },
-    {
-      "epoch": 2.13,
-      "learning_rate": 6.640926640926641e-05,
-      "loss": 0.558,
-      "step": 447
-    },
-    {
-      "epoch": 2.14,
-      "learning_rate": 6.602316602316603e-05,
-      "loss": 0.4858,
-      "step": 448
-    },
-    {
-      "epoch": 2.14,
-      "learning_rate": 6.563706563706564e-05,
-      "loss": 0.6183,
-      "step": 449
-    },
-    {
-      "epoch": 2.15,
-      "learning_rate": 6.525096525096526e-05,
-      "loss": 0.5093,
-      "step": 450
-    },
-    {
-      "epoch": 2.15,
-      "learning_rate": 6.486486486486487e-05,
-      "loss": 0.4336,
-      "step": 451
-    },
-    {
-      "epoch": 2.16,
-      "learning_rate": 6.447876447876449e-05,
-      "loss": 0.653,
-      "step": 452
-    },
-    {
-      "epoch": 2.16,
-      "learning_rate": 6.40926640926641e-05,
-      "loss": 0.5675,
-      "step": 453
-    },
-    {
-      "epoch": 2.17,
-      "learning_rate": 6.37065637065637e-05,
-      "loss": 0.5146,
-      "step": 454
-    },
-    {
-      "epoch": 2.17,
-      "learning_rate": 6.332046332046332e-05,
-      "loss": 0.4988,
-      "step": 455
-    },
-    {
-      "epoch": 2.17,
-      "learning_rate": 6.293436293436293e-05,
-      "loss": 0.5216,
-      "step": 456
-    },
-    {
-      "epoch": 2.18,
-      "learning_rate": 6.254826254826255e-05,
-      "loss": 0.5886,
-      "step": 457
-    },
-    {
-      "epoch": 2.18,
-      "learning_rate": 6.216216216216216e-05,
-      "loss": 0.5856,
-      "step": 458
-    },
-    {
-      "epoch": 2.19,
-      "learning_rate": 6.177606177606178e-05,
-      "loss": 0.4837,
-      "step": 459
-    },
-    {
-      "epoch": 2.19,
-      "learning_rate": 6.138996138996139e-05,
-      "loss": 0.6044,
-      "step": 460
-    },
-    {
-      "epoch": 2.2,
-      "learning_rate": 6.100386100386101e-05,
-      "loss": 0.5276,
-      "step": 461
-    },
-    {
-      "epoch": 2.2,
-      "learning_rate": 6.061776061776062e-05,
-      "loss": 0.4752,
-      "step": 462
-    },
-    {
-      "epoch": 2.21,
-      "learning_rate": 6.023166023166024e-05,
-      "loss": 0.5702,
-      "step": 463
-    },
-    {
-      "epoch": 2.21,
-      "learning_rate": 5.984555984555985e-05,
-      "loss": 0.4758,
-      "step": 464
-    },
-    {
-      "epoch": 2.22,
-      "learning_rate": 5.9459459459459466e-05,
-      "loss": 0.573,
-      "step": 465
-    },
-    {
-      "epoch": 2.22,
-      "eval_loss": 4.131713390350342,
-      "eval_runtime": 7.3741,
-      "eval_samples_per_second": 162.866,
-      "eval_steps_per_second": 54.379,
-      "step": 465
-    },
-    {
-      "epoch": 2.22,
-      "learning_rate": 5.907335907335908e-05,
-      "loss": 0.5373,
-      "step": 466
-    },
-    {
-      "epoch": 2.23,
-      "learning_rate": 5.8687258687258696e-05,
-      "loss": 0.5611,
-      "step": 467
-    },
-    {
-      "epoch": 2.23,
-      "learning_rate": 5.83011583011583e-05,
-      "loss": 0.5744,
-      "step": 468
-    },
-    {
-      "epoch": 2.24,
-      "learning_rate": 5.791505791505791e-05,
-      "loss": 0.4818,
-      "step": 469
-    },
-    {
-      "epoch": 2.24,
-      "learning_rate": 5.752895752895753e-05,
-      "loss": 0.4519,
-      "step": 470
-    },
-    {
-      "epoch": 2.25,
-      "learning_rate": 5.714285714285714e-05,
-      "loss": 0.4295,
-      "step": 471
-    },
-    {
-      "epoch": 2.25,
-      "learning_rate": 5.6756756756756757e-05,
-      "loss": 0.4755,
-      "step": 472
-    },
-    {
-      "epoch": 2.26,
-      "learning_rate": 5.637065637065637e-05,
-      "loss": 0.501,
-      "step": 473
-    },
-    {
-      "epoch": 2.26,
-      "learning_rate": 5.5984555984555986e-05,
-      "loss": 0.449,
-      "step": 474
-    },
-    {
-      "epoch": 2.27,
-      "learning_rate": 5.55984555984556e-05,
-      "loss": 0.4914,
-      "step": 475
-    },
-    {
-      "epoch": 2.27,
-      "learning_rate": 5.5212355212355216e-05,
-      "loss": 0.5153,
-      "step": 476
-    },
-    {
-      "epoch": 2.28,
-      "learning_rate": 5.482625482625483e-05,
-      "loss": 0.5433,
-      "step": 477
-    },
-    {
-      "epoch": 2.28,
-      "learning_rate": 5.4440154440154445e-05,
-      "loss": 0.5248,
-      "step": 478
-    },
-    {
-      "epoch": 2.29,
-      "learning_rate": 5.405405405405406e-05,
-      "loss": 0.5453,
-      "step": 479
-    },
-    {
-      "epoch": 2.29,
-      "learning_rate": 5.3667953667953675e-05,
-      "loss": 0.5288,
-      "step": 480
-    },
-    {
-      "epoch": 2.3,
-      "learning_rate": 5.328185328185329e-05,
-      "loss": 0.532,
-      "step": 481
-    },
-    {
-      "epoch": 2.3,
-      "learning_rate": 5.2895752895752905e-05,
-      "loss": 0.5139,
-      "step": 482
-    },
-    {
-      "epoch": 2.31,
-      "learning_rate": 5.2509652509652506e-05,
-      "loss": 0.5175,
-      "step": 483
-    },
-    {
-      "epoch": 2.31,
-      "learning_rate": 5.212355212355212e-05,
-      "loss": 0.6227,
-      "step": 484
-    },
-    {
-      "epoch": 2.32,
-      "learning_rate": 5.1737451737451736e-05,
-      "loss": 0.567,
-      "step": 485
-    },
-    {
-      "epoch": 2.32,
-      "learning_rate": 5.135135135135135e-05,
-      "loss": 0.5636,
-      "step": 486
-    },
-    {
-      "epoch": 2.33,
-      "learning_rate": 5.0965250965250965e-05,
-      "loss": 0.5367,
-      "step": 487
-    },
-    {
-      "epoch": 2.33,
-      "learning_rate": 5.057915057915058e-05,
-      "loss": 0.6016,
-      "step": 488
-    },
-    {
-      "epoch": 2.33,
-      "learning_rate": 5.0193050193050195e-05,
-      "loss": 0.4492,
-      "step": 489
-    },
-    {
-      "epoch": 2.34,
-      "learning_rate": 4.980694980694981e-05,
-      "loss": 0.5329,
-      "step": 490
-    },
-    {
-      "epoch": 2.34,
-      "learning_rate": 4.9420849420849425e-05,
-      "loss": 0.503,
-      "step": 491
-    },
-    {
-      "epoch": 2.35,
-      "learning_rate": 4.903474903474904e-05,
-      "loss": 0.4799,
-      "step": 492
-    },
-    {
-      "epoch": 2.35,
-      "learning_rate": 4.8648648648648654e-05,
-      "loss": 0.454,
-      "step": 493
-    },
-    {
-      "epoch": 2.36,
-      "learning_rate": 4.826254826254826e-05,
-      "loss": 0.5555,
-      "step": 494
-    },
-    {
-      "epoch": 2.36,
-      "learning_rate": 4.787644787644788e-05,
-      "loss": 0.5925,
-      "step": 495
-    },
-    {
-      "epoch": 2.37,
-      "learning_rate": 4.749034749034749e-05,
-      "loss": 0.5557,
-      "step": 496
-    },
-    {
-      "epoch": 2.37,
-      "eval_loss": 4.199349403381348,
-      "eval_runtime": 7.3742,
-      "eval_samples_per_second": 162.866,
-      "eval_steps_per_second": 54.379,
-      "step": 496
-    },
-    {
-      "epoch": 2.37,
-      "learning_rate": 4.710424710424711e-05,
-      "loss": 0.5157,
-      "step": 497
-    },
-    {
-      "epoch": 2.38,
-      "learning_rate": 4.671814671814672e-05,
-      "loss": 0.5538,
-      "step": 498
-    },
-    {
-      "epoch": 2.38,
-      "learning_rate": 4.6332046332046336e-05,
-      "loss": 0.6174,
-      "step": 499
-    },
-    {
-      "epoch": 2.39,
-      "learning_rate": 4.594594594594595e-05,
-      "loss": 0.4592,
-      "step": 500
-    },
-    {
-      "epoch": 2.39,
-      "learning_rate": 4.555984555984556e-05,
-      "loss": 0.4557,
-      "step": 501
-    },
-    {
-      "epoch": 2.4,
-      "learning_rate": 4.5173745173745174e-05,
-      "loss": 0.5154,
-      "step": 502
-    },
-    {
-      "epoch": 2.4,
-      "learning_rate": 4.478764478764479e-05,
-      "loss": 0.4909,
-      "step": 503
-    },
-    {
-      "epoch": 2.41,
-      "learning_rate": 4.4401544401544404e-05,
-      "loss": 0.4755,
-      "step": 504
-    },
-    {
-      "epoch": 2.41,
-      "learning_rate": 4.401544401544402e-05,
-      "loss": 0.592,
-      "step": 505
-    },
-    {
-      "epoch": 2.42,
-      "learning_rate": 4.3629343629343633e-05,
-      "loss": 0.5014,
-      "step": 506
-    },
-    {
-      "epoch": 2.42,
-      "learning_rate": 4.324324324324325e-05,
-      "loss": 0.4928,
-      "step": 507
-    },
-    {
-      "epoch": 2.43,
-      "learning_rate": 4.2857142857142856e-05,
-      "loss": 0.5352,
-      "step": 508
-    },
-    {
-      "epoch": 2.43,
-      "learning_rate": 4.247104247104247e-05,
-      "loss": 0.5457,
-      "step": 509
-    },
-    {
-      "epoch": 2.44,
-      "learning_rate": 4.2084942084942086e-05,
-      "loss": 0.5182,
-      "step": 510
-    },
-    {
-      "epoch": 2.44,
-      "learning_rate": 4.16988416988417e-05,
-      "loss": 0.527,
-      "step": 511
-    },
-    {
-      "epoch": 2.45,
-      "learning_rate": 4.1312741312741316e-05,
-      "loss": 0.4961,
-      "step": 512
-    },
-    {
-      "epoch": 2.45,
-      "learning_rate": 4.092664092664093e-05,
-      "loss": 0.4988,
-      "step": 513
-    },
-    {
-      "epoch": 2.46,
-      "learning_rate": 4.0540540540540545e-05,
-      "loss": 0.5314,
-      "step": 514
-    },
-    {
-      "epoch": 2.46,
-      "learning_rate": 4.015444015444015e-05,
-      "loss": 0.5523,
-      "step": 515
-    },
-    {
-      "epoch": 2.47,
-      "learning_rate": 3.976833976833977e-05,
-      "loss": 0.4368,
-      "step": 516
-    },
-    {
-      "epoch": 2.47,
-      "learning_rate": 3.938223938223938e-05,
-      "loss": 0.5184,
-      "step": 517
-    },
-    {
-      "epoch": 2.48,
-      "learning_rate": 3.8996138996139e-05,
-      "loss": 0.6171,
-      "step": 518
-    },
-    {
-      "epoch": 2.48,
-      "learning_rate": 3.861003861003861e-05,
-      "loss": 0.5357,
-      "step": 519
-    },
-    {
-      "epoch": 2.49,
-      "learning_rate": 3.822393822393823e-05,
-      "loss": 0.5664,
-      "step": 520
-    },
-    {
-      "epoch": 2.49,
-      "learning_rate": 3.783783783783784e-05,
-      "loss": 0.4805,
-      "step": 521
-    },
-    {
-      "epoch": 2.5,
-      "learning_rate": 3.745173745173745e-05,
-      "loss": 0.4562,
-      "step": 522
-    },
-    {
-      "epoch": 2.5,
-      "learning_rate": 3.7065637065637065e-05,
-      "loss": 0.5238,
-      "step": 523
-    },
-    {
-      "epoch": 2.5,
-      "learning_rate": 3.667953667953668e-05,
-      "loss": 0.4338,
-      "step": 524
-    },
-    {
-      "epoch": 2.51,
-      "learning_rate": 3.6293436293436295e-05,
-      "loss": 0.5656,
-      "step": 525
-    },
-    {
-      "epoch": 2.51,
-      "learning_rate": 3.590733590733591e-05,
-      "loss": 0.4496,
-      "step": 526
-    },
-    {
-      "epoch": 2.52,
-      "learning_rate": 3.5521235521235524e-05,
-      "loss": 0.4997,
-      "step": 527
-    },
-    {
-      "epoch": 2.52,
-      "eval_loss": 4.19807767868042,
-      "eval_runtime": 7.3698,
-      "eval_samples_per_second": 162.962,
-      "eval_steps_per_second": 54.411,
-      "step": 527
-    },
-    {
-      "epoch": 2.52,
-      "learning_rate": 3.513513513513514e-05,
-      "loss": 0.4531,
-      "step": 528
-    },
-    {
-      "epoch": 2.53,
-      "learning_rate": 3.4749034749034754e-05,
-      "loss": 0.5048,
-      "step": 529
-    },
-    {
-      "epoch": 2.53,
-      "learning_rate": 3.436293436293436e-05,
-      "loss": 0.5195,
-      "step": 530
-    },
-    {
-      "epoch": 2.54,
-      "learning_rate": 3.397683397683398e-05,
-      "loss": 0.4885,
-      "step": 531
-    },
-    {
-      "epoch": 2.54,
-      "learning_rate": 3.359073359073359e-05,
-      "loss": 0.6774,
-      "step": 532
-    },
-    {
-      "epoch": 2.55,
-      "learning_rate": 3.3204633204633207e-05,
-      "loss": 0.4755,
-      "step": 533
-    },
-    {
-      "epoch": 2.55,
-      "learning_rate": 3.281853281853282e-05,
-      "loss": 0.5164,
-      "step": 534
-    },
-    {
-      "epoch": 2.56,
-      "learning_rate": 3.2432432432432436e-05,
-      "loss": 0.4748,
-      "step": 535
-    },
-    {
-      "epoch": 2.56,
-      "learning_rate": 3.204633204633205e-05,
-      "loss": 0.5656,
-      "step": 536
-    },
-    {
-      "epoch": 2.57,
-      "learning_rate": 3.166023166023166e-05,
-      "loss": 0.5167,
-      "step": 537
-    },
-    {
-      "epoch": 2.57,
-      "learning_rate": 3.1274131274131274e-05,
-      "loss": 0.5101,
-      "step": 538
-    },
-    {
-      "epoch": 2.58,
-      "learning_rate": 3.088803088803089e-05,
-      "loss": 0.4965,
-      "step": 539
-    },
-    {
-      "epoch": 2.58,
-      "learning_rate": 3.0501930501930504e-05,
-      "loss": 0.5549,
-      "step": 540
-    },
-    {
-      "epoch": 2.59,
-      "learning_rate": 3.011583011583012e-05,
-      "loss": 0.4873,
-      "step": 541
-    },
-    {
-      "epoch": 2.59,
-      "learning_rate": 2.9729729729729733e-05,
-      "loss": 0.5093,
-      "step": 542
-    },
-    {
-      "epoch": 2.6,
-      "learning_rate": 2.9343629343629348e-05,
-      "loss": 0.4897,
-      "step": 543
-    },
-    {
-      "epoch": 2.6,
-      "learning_rate": 2.8957528957528956e-05,
-      "loss": 0.5128,
-      "step": 544
-    },
-    {
-      "epoch": 2.61,
-      "learning_rate": 2.857142857142857e-05,
-      "loss": 0.4829,
-      "step": 545
-    },
-    {
-      "epoch": 2.61,
-      "learning_rate": 2.8185328185328186e-05,
-      "loss": 0.4853,
-      "step": 546
-    },
-    {
-      "epoch": 2.62,
-      "learning_rate": 2.77992277992278e-05,
-      "loss": 0.5499,
-      "step": 547
-    },
-    {
-      "epoch": 2.62,
-      "learning_rate": 2.7413127413127415e-05,
-      "loss": 0.59,
-      "step": 548
-    },
-    {
-      "epoch": 2.63,
-      "learning_rate": 2.702702702702703e-05,
-      "loss": 0.5169,
-      "step": 549
-    },
-    {
-      "epoch": 2.63,
-      "learning_rate": 2.6640926640926645e-05,
-      "loss": 0.5642,
-      "step": 550
-    },
-    {
-      "epoch": 2.64,
-      "learning_rate": 2.6254826254826253e-05,
-      "loss": 0.5745,
-      "step": 551
-    },
-    {
-      "epoch": 2.64,
-      "learning_rate": 2.5868725868725868e-05,
-      "loss": 0.4845,
-      "step": 552
-    },
-    {
-      "epoch": 2.65,
-      "learning_rate": 2.5482625482625483e-05,
-      "loss": 0.5198,
-      "step": 553
-    },
-    {
-      "epoch": 2.65,
-      "learning_rate": 2.5096525096525097e-05,
-      "loss": 0.5402,
-      "step": 554
-    },
-    {
-      "epoch": 2.66,
-      "learning_rate": 2.4710424710424712e-05,
-      "loss": 0.5122,
-      "step": 555
-    },
-    {
-      "epoch": 2.66,
-      "learning_rate": 2.4324324324324327e-05,
-      "loss": 0.5769,
-      "step": 556
-    },
-    {
-      "epoch": 2.67,
-      "learning_rate": 2.393822393822394e-05,
-      "loss": 0.5519,
-      "step": 557
-    },
-    {
-      "epoch": 2.67,
-      "learning_rate": 2.3552123552123553e-05,
-      "loss": 0.465,
-      "step": 558
-    },
-    {
-      "epoch": 2.67,
-      "eval_loss": 4.177217483520508,
-      "eval_runtime": 7.3742,
-      "eval_samples_per_second": 162.865,
-      "eval_steps_per_second": 54.379,
-      "step": 558
-    },
-    {
-      "epoch": 2.67,
-      "learning_rate": 2.3166023166023168e-05,
-      "loss": 0.4816,
-      "step": 559
-    },
-    {
-      "epoch": 2.68,
-      "learning_rate": 2.277992277992278e-05,
-      "loss": 0.4428,
-      "step": 560
-    },
-    {
-      "epoch": 2.68,
-      "learning_rate": 2.2393822393822394e-05,
-      "loss": 0.4969,
-      "step": 561
-    },
-    {
-      "epoch": 2.69,
-      "learning_rate": 2.200772200772201e-05,
-      "loss": 0.4891,
-      "step": 562
-    },
-    {
-      "epoch": 2.69,
-      "learning_rate": 2.1621621621621624e-05,
-      "loss": 0.4082,
-      "step": 563
-    },
-    {
-      "epoch": 2.7,
-      "learning_rate": 2.1235521235521236e-05,
-      "loss": 0.4735,
-      "step": 564
-    },
-    {
-      "epoch": 2.7,
-      "learning_rate": 2.084942084942085e-05,
-      "loss": 0.5121,
-      "step": 565
-    },
-    {
-      "epoch": 2.71,
-      "learning_rate": 2.0463320463320465e-05,
-      "loss": 0.4696,
-      "step": 566
-    },
-    {
-      "epoch": 2.71,
-      "learning_rate": 2.0077220077220077e-05,
-      "loss": 0.397,
-      "step": 567
-    },
-    {
-      "epoch": 2.72,
-      "learning_rate": 1.969111969111969e-05,
-      "loss": 0.5271,
-      "step": 568
-    },
-    {
-      "epoch": 2.72,
-      "learning_rate": 1.9305019305019306e-05,
-      "loss": 0.4974,
-      "step": 569
-    },
-    {
-      "epoch": 2.73,
-      "learning_rate": 1.891891891891892e-05,
-      "loss": 0.4814,
-      "step": 570
-    },
-    {
-      "epoch": 2.73,
-      "learning_rate": 1.8532818532818533e-05,
-      "loss": 0.5565,
-      "step": 571
-    },
-    {
-      "epoch": 2.74,
-      "learning_rate": 1.8146718146718147e-05,
-      "loss": 0.4737,
-      "step": 572
-    },
-    {
-      "epoch": 2.74,
-      "learning_rate": 1.7760617760617762e-05,
-      "loss": 0.4448,
-      "step": 573
-    },
-    {
-      "epoch": 2.75,
-      "learning_rate": 1.7374517374517377e-05,
-      "loss": 0.4886,
-      "step": 574
-    },
-    {
-      "epoch": 2.75,
-      "learning_rate": 1.698841698841699e-05,
-      "loss": 0.5197,
-      "step": 575
-    },
-    {
-      "epoch": 2.76,
-      "learning_rate": 1.6602316602316603e-05,
-      "loss": 0.4688,
-      "step": 576
-    },
-    {
-      "epoch": 2.76,
-      "learning_rate": 1.6216216216216218e-05,
-      "loss": 0.5649,
-      "step": 577
-    },
-    {
-      "epoch": 2.77,
-      "learning_rate": 1.583011583011583e-05,
-      "loss": 0.5026,
-      "step": 578
-    },
-    {
-      "epoch": 2.77,
-      "learning_rate": 1.5444015444015444e-05,
-      "loss": 0.5832,
-      "step": 579
-    },
-    {
-      "epoch": 2.78,
-      "learning_rate": 1.505791505791506e-05,
-      "loss": 0.5995,
-      "step": 580
-    },
-    {
-      "epoch": 2.78,
-      "learning_rate": 1.4671814671814674e-05,
-      "loss": 0.5342,
-      "step": 581
-    },
-    {
-      "epoch": 2.79,
-      "learning_rate": 1.4285714285714285e-05,
-      "loss": 0.5465,
-      "step": 582
-    },
-    {
-      "epoch": 2.79,
-      "learning_rate": 1.38996138996139e-05,
-      "loss": 0.5165,
-      "step": 583
-    },
-    {
-      "epoch": 2.8,
-      "learning_rate": 1.3513513513513515e-05,
-      "loss": 0.4594,
-      "step": 584
-    },
-    {
-      "epoch": 2.8,
-      "learning_rate": 1.3127413127413127e-05,
-      "loss": 0.4448,
-      "step": 585
-    },
-    {
-      "epoch": 2.81,
-      "learning_rate": 1.2741312741312741e-05,
-      "loss": 0.5148,
-      "step": 586
-    },
-    {
-      "epoch": 2.81,
-      "learning_rate": 1.2355212355212356e-05,
-      "loss": 0.5255,
-      "step": 587
-    },
-    {
-      "epoch": 2.82,
-      "learning_rate": 1.196911196911197e-05,
-      "loss": 0.4979,
-      "step": 588
-    },
-    {
-      "epoch": 2.82,
-      "learning_rate": 1.1583011583011584e-05,
-      "loss": 0.4531,
-      "step": 589
-    },
-    {
-      "epoch": 2.82,
-      "eval_loss": 4.171577453613281,
-      "eval_runtime": 7.3719,
-      "eval_samples_per_second": 162.916,
-      "eval_steps_per_second": 54.396,
-      "step": 589
-    },
-    {
-      "epoch": 2.83,
-      "learning_rate": 1.1196911196911197e-05,
-      "loss": 0.5339,
-      "step": 590
-    },
-    {
-      "epoch": 2.83,
-      "learning_rate": 1.0810810810810812e-05,
-      "loss": 0.5242,
-      "step": 591
-    },
-    {
-      "epoch": 2.83,
-      "learning_rate": 1.0424710424710425e-05,
-      "loss": 0.5266,
-      "step": 592
-    },
-    {
-      "epoch": 2.84,
-      "learning_rate": 1.0038610038610038e-05,
-      "loss": 0.5188,
-      "step": 593
-    },
-    {
-      "epoch": 2.84,
-      "learning_rate": 9.652509652509653e-06,
-      "loss": 0.459,
-      "step": 594
-    },
-    {
-      "epoch": 2.85,
-      "learning_rate": 9.266409266409266e-06,
-      "loss": 0.3489,
-      "step": 595
-    },
-    {
-      "epoch": 2.85,
-      "learning_rate": 8.880308880308881e-06,
-      "loss": 0.5022,
-      "step": 596
-    },
-    {
-      "epoch": 2.86,
-      "learning_rate": 8.494208494208494e-06,
-      "loss": 0.4513,
-      "step": 597
-    },
-    {
-      "epoch": 2.86,
-      "learning_rate": 8.108108108108109e-06,
-      "loss": 0.4338,
-      "step": 598
-    },
-    {
-      "epoch": 2.87,
-      "learning_rate": 7.722007722007722e-06,
-      "loss": 0.5263,
-      "step": 599
-    },
-    {
-      "epoch": 2.87,
-      "learning_rate": 7.335907335907337e-06,
-      "loss": 0.4898,
-      "step": 600
-    },
-    {
-      "epoch": 2.88,
-      "learning_rate": 6.94980694980695e-06,
-      "loss": 0.5212,
-      "step": 601
-    },
-    {
-      "epoch": 2.88,
-      "learning_rate": 6.563706563706563e-06,
-      "loss": 0.5355,
-      "step": 602
-    },
-    {
-      "epoch": 2.89,
-      "learning_rate": 6.177606177606178e-06,
-      "loss": 0.5444,
-      "step": 603
-    },
-    {
-      "epoch": 2.89,
-      "learning_rate": 5.791505791505792e-06,
-      "loss": 0.5875,
-      "step": 604
-    },
-    {
-      "epoch": 2.9,
-      "learning_rate": 5.405405405405406e-06,
-      "loss": 0.4989,
-      "step": 605
-    },
-    {
-      "epoch": 2.9,
-      "learning_rate": 5.019305019305019e-06,
-      "loss": 0.5159,
-      "step": 606
-    },
-    {
-      "epoch": 2.91,
-      "learning_rate": 4.633204633204633e-06,
-      "loss": 0.3956,
-      "step": 607
-    },
-    {
-      "epoch": 2.91,
-      "learning_rate": 4.247104247104247e-06,
-      "loss": 0.491,
-      "step": 608
-    },
-    {
-      "epoch": 2.92,
-      "learning_rate": 3.861003861003861e-06,
-      "loss": 0.4454,
-      "step": 609
-    },
-    {
-      "epoch": 2.92,
-      "learning_rate": 3.474903474903475e-06,
-      "loss": 0.4844,
-      "step": 610
-    },
-    {
-      "epoch": 2.93,
-      "learning_rate": 3.088803088803089e-06,
-      "loss": 0.4972,
-      "step": 611
-    },
-    {
-      "epoch": 2.93,
-      "learning_rate": 2.702702702702703e-06,
-      "loss": 0.4617,
-      "step": 612
-    },
-    {
-      "epoch": 2.94,
-      "learning_rate": 2.3166023166023166e-06,
-      "loss": 0.4683,
-      "step": 613
-    },
-    {
-      "epoch": 2.94,
-      "learning_rate": 1.9305019305019305e-06,
-      "loss": 0.5815,
-      "step": 614
-    },
-    {
-      "epoch": 2.95,
-      "learning_rate": 1.5444015444015445e-06,
-      "loss": 0.4502,
-      "step": 615
-    },
-    {
-      "epoch": 2.95,
-      "learning_rate": 1.1583011583011583e-06,
-      "loss": 0.477,
-      "step": 616
-    },
-    {
-      "epoch": 2.96,
-      "learning_rate": 7.722007722007723e-07,
-      "loss": 0.4836,
-      "step": 617
-    },
-    {
-      "epoch": 2.96,
-      "learning_rate": 3.8610038610038613e-07,
-      "loss": 0.4523,
-      "step": 618
-    }
-  ],
-  "logging_steps": 1,
-  "max_steps": 618,
-  "num_train_epochs": 3,
-  "save_steps": 500,
-  "total_flos": 3.072196311043277e+17,
-  "trial_name": null,
-  "trial_params": null
-}

checkpoint-618/training_args.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:40caa3ffe88e39fb8f17ca2f4b2952df2344fe3de9435a3b5cb8662a65ff745d
-size 6011

checkpoint-618/zero_to_fp32.py

@@ -1,587 +0,0 @@
-#!/usr/bin/env python
-
-# Copyright (c) Microsoft Corporation.
-# SPDX-License-Identifier: Apache-2.0
-
-# DeepSpeed Team
-
-# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
-# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
-# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
-# application.
-#
-# example: python zero_to_fp32.py . pytorch_model.bin
-
-import argparse
-import torch
-import glob
-import math
-import os
-import re
-from collections import OrderedDict
-from dataclasses import dataclass
-
-# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
-# DeepSpeed data structures it has to be available in the current python environment.
-from deepspeed.utils import logger
-from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
-                                            FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
-                                            FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
-
-
-@dataclass
-class zero_model_state:
-    buffers: dict()
-    param_shapes: dict()
-    shared_params: list
-    ds_version: int
-    frozen_param_shapes: dict()
-    frozen_param_fragments: dict()
-
-
-debug = 0
-
-# load to cpu
-device = torch.device('cpu')
-
-
-def atoi(text):
-    return int(text) if text.isdigit() else text
-
-
-def natural_keys(text):
-    '''
-    alist.sort(key=natural_keys) sorts in human order
-    http://nedbatchelder.com/blog/200712/human_sorting.html
-    (See Toothy's implementation in the comments)
-    '''
-    return [atoi(c) for c in re.split(r'(\d+)', text)]
-
-
-def get_model_state_file(checkpoint_dir, zero_stage):
-    if not os.path.isdir(checkpoint_dir):
-        raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
-
-    # there should be only one file
-    if zero_stage <= 2:
-        file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
-    elif zero_stage == 3:
-        file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
-
-    if not os.path.exists(file):
-        raise FileNotFoundError(f"can't find model states file at '{file}'")
-
-    return file
-
-
-def get_checkpoint_files(checkpoint_dir, glob_pattern):
-    # XXX: need to test that this simple glob rule works for multi-node setup too
-    ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
-
-    if len(ckpt_files) == 0:
-        raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
-
-    return ckpt_files
-
-
-def get_optim_files(checkpoint_dir):
-    return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
-
-
-def get_model_state_files(checkpoint_dir):
-    return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
-
-
-def parse_model_states(files):
-    zero_model_states = []
-    for file in files:
-        state_dict = torch.load(file, map_location=device)
-
-        if BUFFER_NAMES not in state_dict:
-            raise ValueError(f"{file} is not a model state checkpoint")
-        buffer_names = state_dict[BUFFER_NAMES]
-        if debug:
-            print("Found buffers:", buffer_names)
-
-        # recover just the buffers while restoring them to fp32 if they were saved in fp16
-        buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
-        param_shapes = state_dict[PARAM_SHAPES]
-
-        # collect parameters that are included in param_shapes
-        param_names = []
-        for s in param_shapes:
-            for name in s.keys():
-                param_names.append(name)
-
-        # update with frozen parameters
-        frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
-        if frozen_param_shapes is not None:
-            if debug:
-                print(f"Found frozen_param_shapes: {frozen_param_shapes}")
-            param_names += list(frozen_param_shapes.keys())
-
-        # handle shared params
-        shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
-
-        ds_version = state_dict.get(DS_VERSION, None)
-
-        frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
-
-        z_model_state = zero_model_state(buffers=buffers,
-                                         param_shapes=param_shapes,
-                                         shared_params=shared_params,
-                                         ds_version=ds_version,
-                                         frozen_param_shapes=frozen_param_shapes,
-                                         frozen_param_fragments=frozen_param_fragments)
-        zero_model_states.append(z_model_state)
-
-    return zero_model_states
-
-
-def parse_optim_states(files, ds_checkpoint_dir):
-
-    total_files = len(files)
-    state_dicts = []
-    for f in files:
-        state_dict = torch.load(f, map_location=device)
-        # immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
-        # and also handle the case where it was already removed by another helper script
-        state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
-        state_dicts.append(state_dict)
-
-    if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
-        raise ValueError(f"{files[0]} is not a zero checkpoint")
-    zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
-    world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
-
-    # For ZeRO-2 each param group can have different partition_count as data parallelism for expert
-    # parameters can be different from data parallelism for non-expert parameters. So we can just
-    # use the max of the partition_count to get the dp world_size.
-
-    if type(world_size) is list:
-        world_size = max(world_size)
-
-    if world_size != total_files:
-        raise ValueError(
-            f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
-            "Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
-        )
-
-    # the groups are named differently in each stage
-    if zero_stage <= 2:
-        fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
-    elif zero_stage == 3:
-        fp32_groups_key = FP32_FLAT_GROUPS
-    else:
-        raise ValueError(f"unknown zero stage {zero_stage}")
-
-    if zero_stage <= 2:
-        fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
-    elif zero_stage == 3:
-        # if there is more than one param group, there will be multiple flattened tensors - one
-        # flattened tensor per group - for simplicity merge them into a single tensor
-        #
-        # XXX: could make the script more memory efficient for when there are multiple groups - it
-        # will require matching the sub-lists of param_shapes for each param group flattened tensor
-
-        fp32_flat_groups = [
-            torch.cat(state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key], 0) for i in range(len(state_dicts))
-        ]
-
-    return zero_stage, world_size, fp32_flat_groups
-
-
-def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir):
-    """
-    Returns fp32 state_dict reconstructed from ds checkpoint
-
-    Args:
-        - ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
-
-    """
-    print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
-
-    optim_files = get_optim_files(ds_checkpoint_dir)
-    zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
-    print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
-
-    model_files = get_model_state_files(ds_checkpoint_dir)
-
-    zero_model_states = parse_model_states(model_files)
-    print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
-
-    if zero_stage <= 2:
-        return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states)
-    elif zero_stage == 3:
-        return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states)
-
-
-def _zero2_merge_frozen_params(state_dict, zero_model_states):
-    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
-        return
-
-    frozen_param_shapes = zero_model_states[0].frozen_param_shapes
-    frozen_param_fragments = zero_model_states[0].frozen_param_fragments
-
-    if debug:
-        num_elem = sum(s.numel() for s in frozen_param_shapes.values())
-        print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
-
-        wanted_params = len(frozen_param_shapes)
-        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
-        avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
-        print(f'Frozen params: Have {avail_numel} numels to process.')
-        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
-
-    total_params = 0
-    total_numel = 0
-    for name, shape in frozen_param_shapes.items():
-        total_params += 1
-        unpartitioned_numel = shape.numel()
-        total_numel += unpartitioned_numel
-
-        state_dict[name] = frozen_param_fragments[name]
-
-        if debug:
-            print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
-
-    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
-
-
-def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
-    param_shapes = zero_model_states[0].param_shapes
-
-    # Reconstruction protocol:
-    #
-    # XXX: document this
-
-    if debug:
-        for i in range(world_size):
-            for j in range(len(fp32_flat_groups[0])):
-                print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
-
-    # XXX: memory usage doubles here (zero2)
-    num_param_groups = len(fp32_flat_groups[0])
-    merged_single_partition_of_fp32_groups = []
-    for i in range(num_param_groups):
-        merged_partitions = [sd[i] for sd in fp32_flat_groups]
-        full_single_fp32_vector = torch.cat(merged_partitions, 0)
-        merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
-    avail_numel = sum(
-        [full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
-
-    if debug:
-        wanted_params = sum([len(shapes) for shapes in param_shapes])
-        wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
-        # not asserting if there is a mismatch due to possible padding
-        print(f"Have {avail_numel} numels to process.")
-        print(f"Need {wanted_numel} numels in {wanted_params} params.")
-
-    # params
-    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
-    # out-of-core computing solution
-    total_numel = 0
-    total_params = 0
-    for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
-        offset = 0
-        avail_numel = full_single_fp32_vector.numel()
-        for name, shape in shapes.items():
-
-            unpartitioned_numel = shape.numel()
-            total_numel += unpartitioned_numel
-            total_params += 1
-
-            if debug:
-                print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
-            state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
-            offset += unpartitioned_numel
-
-        # Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
-        # avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
-        # paddings performed in the code it's almost impossible to predict the exact numbers w/o the
-        # live optimizer object, so we are checking that the numbers are within the right range
-        align_to = 2 * world_size
-
-        def zero2_align(x):
-            return align_to * math.ceil(x / align_to)
-
-        if debug:
-            print(f"original offset={offset}, avail_numel={avail_numel}")
-
-        offset = zero2_align(offset)
-        avail_numel = zero2_align(avail_numel)
-
-        if debug:
-            print(f"aligned  offset={offset}, avail_numel={avail_numel}")
-
-        # Sanity check
-        if offset != avail_numel:
-            raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
-
-    print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
-
-
-def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states):
-    state_dict = OrderedDict()
-
-    # buffers
-    buffers = zero_model_states[0].buffers
-    state_dict.update(buffers)
-    if debug:
-        print(f"added {len(buffers)} buffers")
-
-    _zero2_merge_frozen_params(state_dict, zero_model_states)
-
-    _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
-
-    # recover shared parameters
-    for pair in zero_model_states[0].shared_params:
-        if pair[1] in state_dict:
-            state_dict[pair[0]] = state_dict[pair[1]]
-
-    return state_dict
-
-
-def zero3_partitioned_param_info(unpartitioned_numel, world_size):
-    remainder = unpartitioned_numel % world_size
-    padding_numel = (world_size - remainder) if remainder else 0
-    partitioned_numel = math.ceil(unpartitioned_numel / world_size)
-    return partitioned_numel, padding_numel
-
-
-def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
-    if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
-        return
-
-    if debug:
-        for i in range(world_size):
-            num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
-            print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
-
-        frozen_param_shapes = zero_model_states[0].frozen_param_shapes
-        wanted_params = len(frozen_param_shapes)
-        wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
-        avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
-        print(f'Frozen params: Have {avail_numel} numels to process.')
-        print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
-
-    total_params = 0
-    total_numel = 0
-    for name, shape in zero_model_states[0].frozen_param_shapes.items():
-        total_params += 1
-        unpartitioned_numel = shape.numel()
-        total_numel += unpartitioned_numel
-
-        param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
-        state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
-
-        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
-
-        if debug:
-            print(
-                f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
-            )
-
-    print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
-
-
-def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
-    param_shapes = zero_model_states[0].param_shapes
-    avail_numel = fp32_flat_groups[0].numel() * world_size
-    # Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
-    # param, re-consolidating each param, while dealing with padding if any
-
-    # merge list of dicts, preserving order
-    param_shapes = {k: v for d in param_shapes for k, v in d.items()}
-
-    if debug:
-        for i in range(world_size):
-            print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
-
-        wanted_params = len(param_shapes)
-        wanted_numel = sum(shape.numel() for shape in param_shapes.values())
-        # not asserting if there is a mismatch due to possible padding
-        avail_numel = fp32_flat_groups[0].numel() * world_size
-        print(f"Trainable params: Have {avail_numel} numels to process.")
-        print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
-
-    # params
-    # XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
-    # out-of-core computing solution
-    offset = 0
-    total_numel = 0
-    total_params = 0
-    for name, shape in param_shapes.items():
-
-        unpartitioned_numel = shape.numel()
-        total_numel += unpartitioned_numel
-        total_params += 1
-
-        partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
-
-        if debug:
-            print(
-                f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
-            )
-
-        # XXX: memory usage doubles here
-        state_dict[name] = torch.cat(
-            tuple(fp32_flat_groups[i].narrow(0, offset, partitioned_numel) for i in range(world_size)),
-            0).narrow(0, 0, unpartitioned_numel).view(shape)
-        offset += partitioned_numel
-
-    offset *= world_size
-
-    # Sanity check
-    if offset != avail_numel:
-        raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
-
-    print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
-
-
-def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states):
-    state_dict = OrderedDict()
-
-    # buffers
-    buffers = zero_model_states[0].buffers
-    state_dict.update(buffers)
-    if debug:
-        print(f"added {len(buffers)} buffers")
-
-    _zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
-
-    _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
-
-    # recover shared parameters
-    for pair in zero_model_states[0].shared_params:
-        if pair[1] in state_dict:
-            state_dict[pair[0]] = state_dict[pair[1]]
-
-    return state_dict
-
-
-def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag=None):
-    """
-    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
-    ``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
-    via a model hub.
-
-    Args:
-        - ``checkpoint_dir``: path to the desired checkpoint folder
-        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
-
-    Returns:
-        - pytorch ``state_dict``
-
-    Note: this approach may not work if your application doesn't have sufficient free CPU memory and
-    you may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
-    the checkpoint.
-
-    A typical usage might be ::
-
-        from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
-        # do the training and checkpoint saving
-        state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
-        model = model.cpu() # move to cpu
-        model.load_state_dict(state_dict)
-        # submit to model hub or save the model to share with others
-
-    In this example the ``model`` will no longer be usable in the deepspeed context of the same
-    application. i.e. you will need to re-initialize the deepspeed engine, since
-    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
-
-    If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
-
-    """
-    if tag is None:
-        latest_path = os.path.join(checkpoint_dir, 'latest')
-        if os.path.isfile(latest_path):
-            with open(latest_path, 'r') as fd:
-                tag = fd.read().strip()
-        else:
-            raise ValueError(f"Unable to find 'latest' file at {latest_path}")
-
-    ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
-
-    if not os.path.isdir(ds_checkpoint_dir):
-        raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
-
-    return _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir)
-
-
-def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir, output_file, tag=None):
-    """
-    Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
-    loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
-
-    Args:
-        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
-        - ``output_file``: path to the pytorch fp32 state_dict output file (e.g. path/pytorch_model.bin)
-        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
-    """
-
-    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
-    print(f"Saving fp32 state dict to {output_file}")
-    torch.save(state_dict, output_file)
-
-
-def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
-    """
-    1. Put the provided model to cpu
-    2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
-    3. Load it into the provided model
-
-    Args:
-        - ``model``: the model object to update
-        - ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
-        - ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
-
-    Returns:
-        - ``model`: modified model
-
-    Make sure you have plenty of CPU memory available before you call this function. If you don't
-    have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
-    conveniently placed for you in the checkpoint folder.
-
-    A typical usage might be ::
-
-        from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
-        model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
-        # submit to model hub or save the model to share with others
-
-    Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
-    of the same application. i.e. you will need to re-initialize the deepspeed engine, since
-    ``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
-
-    """
-    logger.info(f"Extracting fp32 weights")
-    state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
-
-    logger.info(f"Overwriting model with fp32 weights")
-    model = model.cpu()
-    model.load_state_dict(state_dict, strict=False)
-
-    return model
-
-
-if __name__ == "__main__":
-
-    parser = argparse.ArgumentParser()
-    parser.add_argument("checkpoint_dir",
-                        type=str,
-                        help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
-    parser.add_argument(
-        "output_file",
-        type=str,
-        help="path to the pytorch fp32 state_dict output file (e.g. path/checkpoint-12/pytorch_model.bin)")
-    parser.add_argument("-t",
-                        "--tag",
-                        type=str,
-                        default=None,
-                        help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
-    parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
-    args = parser.parse_args()
-
-    debug = args.debug
-
-    convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir, args.output_file, tag=args.tag)