update model

config.json

@@ -0,0 +1,24 @@
+{
+  "_name_or_path": "decapoda-research/llama-7b-hf",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "bos_token_id": 0,
+  "eos_token_id": 1,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_position_embeddings": 2048,
+  "max_sequence_length": 2048,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "pad_token_id": -1,
+  "rms_norm_eps": 1e-06,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.30.0.dev0",
+  "use_cache": true,
+  "vocab_size": 32000
+}

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 0,
+  "eos_token_id": 1,
+  "pad_token_id": 0,
+  "transformers_version": "4.30.0.dev0"
+}

pytorch_model-00001-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a7ca70d138b14570b5f8752df70a65c4b033992fab11dffaca50b7cff7aac47
+size 396364479

pytorch_model-00002-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6163f92fd79f93dd20c8c3004e33dbc8f2ce5d24e889fd92708d550de0e06c7
+size 371215393

pytorch_model-00003-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0de32c6534b4f0bc04ea4e000c1b84cdcf5842b0e340e37396c75ba307c8b4b
+size 371215986

pytorch_model-00004-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd2e6d90ff944cf8e564aeb8e51f1a6ad2dc348185f0133c20756cfcfe1f8fd7
+size 371215986

pytorch_model-00005-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:778e2a2205118e9e5a9b32a95d5a4f7d34304615f17a3d3d4fbe284980d53e45
+size 371215986

pytorch_model-00006-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f7367411aa541b1dbb642928961efccd0b2b984ffde8124ae2b125d9d6488b1
+size 314575888

pytorch_model-00007-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21c8673bb05ab8e56f9fc0756b415eacca8df41de84fad4b79856edef94f8ca2
+size 314592882

pytorch_model-00008-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:124c6b76ac1a4a495a8b238b1c9f41d89a27bf2a2daa3c8489bd857b59552642
+size 314592882

pytorch_model-00009-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70e73295a2c47eb6aa19cbf7ef6bd5ea96ac55f00173039769172f499def3ccb
+size 371215393

pytorch_model-00010-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:962772c57b9687947e039eb07767959c3ce4d452434149a5f32b0c7a0285ed0d
+size 371215986

pytorch_model-00011-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8261e481ca628e5f00ede0bff87f5e5f9e87d8d715e9b09a3572a66164c69932
+size 371215986

pytorch_model-00012-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a9a83838fbf6dbf2aae017aabcbf178dfca8938298afbe26f9f517ada161db1
+size 371215986

pytorch_model-00013-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c903e4a4d9881205c46b4abd86929521a891398930a6789202ecbd19bd1129c
+size 314575888

pytorch_model-00014-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3249582444b31bf1c56aa4858f4cd6dc37b3a83dfa45766f5f8ca5616c0c227e
+size 314592882

pytorch_model-00015-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:05679259916f0cd6824c21d72e6193462c6a96c56273a78398a77684464c7ac3
+size 314592882

pytorch_model-00016-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57c6781059ed1f50e7d780218fa2479c6970ef38ce16356ace4170c4ea153c5f
+size 371215393

pytorch_model-00017-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:093b9d9cccdf4d15bb10ed9de6ddc009c27d749dabbe0a1e4ef9152c57bd679d
+size 371215986

pytorch_model-00018-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba618510ce45cbd188a5c7c22e6bcf501b21264b4e6e566dccf118647dc84b67
+size 371215986

pytorch_model-00019-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4e2e7e5241f9b198851a07a0a208101cebe7750c2c6e1f7c024bed1480f32d18
+size 371215986

pytorch_model-00020-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f09319b8cf960aa5f4a21f2b83edbcf334d7e7b91deff98b111ae559090eafce
+size 314575888

pytorch_model-00021-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8038cd437371122d433e439c9cd04964d1516e3c74708abb1f08929c50c31c78
+size 314592882

pytorch_model-00022-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a43c5a1ab85616a31df96866ba926e17a48eec015dc528d38b82ac9dcd34e65
+size 314592882

pytorch_model-00023-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce991769b8b59a8cb6a52827f5129696263896928ab8bd56c9d0ec20fc03ac9b
+size 371215393

pytorch_model-00024-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f2ef305c7027ee65072bc8424fd7252ff4956ce233ddfa5c5cc6b3d14964fee
+size 371215986

pytorch_model-00025-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9364e17a4d69f40f05d5cc5360ad5f1ffb47752f736e571db7658732f9c4b62
+size 371215986

pytorch_model-00026-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9926cb2a5a7f0dc866f2a8fab75659a3066707a18a55e21d9821a1ad5ac0e594
+size 371215986

pytorch_model-00027-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87b18eafd142cefd70e6be5ed71f08cce65bce04fda77b9f728bf94436fee4ac
+size 314575888

pytorch_model-00028-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d66180b7dfdca59c1c72f604554b376ab1f95bce8b57246f12a3a6bfafa1335
+size 314592882

pytorch_model-00029-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:13b75b8d09cc4e416fc82a1ce013eeda27c8233699c272e30f4e6392da43ce32
+size 314592882

pytorch_model-00030-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:388916bc4af3187b4669422cffd778b2cabc588ee18cb93c61dad37abcf7255d
+size 371215393

pytorch_model-00031-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a95c0c95e94edf5429e89bd66107568cd8a0f91671e6c72c6fc11cf6c4a7d475
+size 371215986

pytorch_model-00032-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62ec94bbe73d0e03b24005db06d9b61479e28c8edc96369c771439e763e4d503
+size 371215986

pytorch_model-00033-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc5086f9fdd4a76ad5926949a18a00252401c610b34e2502cacba3d7fd286784
+size 371215986

pytorch_model-00034-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8640cda463ca9a3dea54438250c4d0febf755dd8573c25bcf946413cd4826d1a
+size 314575888

pytorch_model-00035-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:66077b63c5f0e892b023db2b9dcd14ccabd27085a23a61b6356172f82c0751e7
+size 314592882

pytorch_model-00036-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c50669dec7327f6741304af8f0e333632cbc71dd94fd8e8847b3f64e4baee63
+size 314592882

pytorch_model-00037-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a07e84a9856a3d06085a48a49a8804f8bcc7f6a5bb26bde0e47ce9dc7de68cae
+size 371215393

pytorch_model-00038-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d212079f2e1e1070bac9d22fb87736e78550959b30496072c04c18bc95568a30
+size 304114977

pytorch_model-00039-of-00039.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3350fb71c2bc1e88d77d9685716fd2759e879137814a4120d98265a43f405919
+size 262144938

pytorch_model.bin.index.json

@@ -0,0 +1,330 @@
+{
+  "metadata": {
+    "total_size": 13476839424
+  },
+  "weight_map": {
+    "lm_head.weight": "pytorch_model-00039-of-00039.bin",
+    "model.embed_tokens.weight": "pytorch_model-00001-of-00039.bin",
+    "model.layers.0.input_layernorm.weight": "pytorch_model-00002-of-00039.bin",
+    "model.layers.0.mlp.down_proj.weight": "pytorch_model-00002-of-00039.bin",
+    "model.layers.0.mlp.gate_proj.weight": "pytorch_model-00002-of-00039.bin",
+    "model.layers.0.mlp.up_proj.weight": "pytorch_model-00002-of-00039.bin",
+    "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00002-of-00039.bin",
+    "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00039.bin",
+    "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00039.bin",
+    "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00039.bin",
+    "model.layers.0.self_attn.rotary_emb.inv_freq": "pytorch_model-00001-of-00039.bin",
+    "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00039.bin",
+    "model.layers.1.input_layernorm.weight": "pytorch_model-00003-of-00039.bin",
+    "model.layers.1.mlp.down_proj.weight": "pytorch_model-00003-of-00039.bin",
+    "model.layers.1.mlp.gate_proj.weight": "pytorch_model-00003-of-00039.bin",
+    "model.layers.1.mlp.up_proj.weight": "pytorch_model-00003-of-00039.bin",
+    "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00003-of-00039.bin",
+    "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00002-of-00039.bin",
+    "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00003-of-00039.bin",
+    "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00002-of-00039.bin",
+    "model.layers.1.self_attn.rotary_emb.inv_freq": "pytorch_model-00003-of-00039.bin",
+    "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00002-of-00039.bin",
+    "model.layers.10.input_layernorm.weight": "pytorch_model-00014-of-00039.bin",
+    "model.layers.10.mlp.down_proj.weight": "pytorch_model-00013-of-00039.bin",
+    "model.layers.10.mlp.gate_proj.weight": "pytorch_model-00013-of-00039.bin",
+    "model.layers.10.mlp.up_proj.weight": "pytorch_model-00014-of-00039.bin",
+    "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00014-of-00039.bin",
+    "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00013-of-00039.bin",
+    "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00013-of-00039.bin",
+    "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00013-of-00039.bin",
+    "model.layers.10.self_attn.rotary_emb.inv_freq": "pytorch_model-00013-of-00039.bin",
+    "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00013-of-00039.bin",
+    "model.layers.11.input_layernorm.weight": "pytorch_model-00015-of-00039.bin",
+    "model.layers.11.mlp.down_proj.weight": "pytorch_model-00015-of-00039.bin",
+    "model.layers.11.mlp.gate_proj.weight": "pytorch_model-00014-of-00039.bin",
+    "model.layers.11.mlp.up_proj.weight": "pytorch_model-00015-of-00039.bin",
+    "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00015-of-00039.bin",
+    "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00014-of-00039.bin",
+    "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00014-of-00039.bin",
+    "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00014-of-00039.bin",
+    "model.layers.11.self_attn.rotary_emb.inv_freq": "pytorch_model-00014-of-00039.bin",
+    "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00014-of-00039.bin",
+    "model.layers.12.input_layernorm.weight": "pytorch_model-00016-of-00039.bin",
+    "model.layers.12.mlp.down_proj.weight": "pytorch_model-00016-of-00039.bin",
+    "model.layers.12.mlp.gate_proj.weight": "pytorch_model-00016-of-00039.bin",
+    "model.layers.12.mlp.up_proj.weight": "pytorch_model-00016-of-00039.bin",
+    "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00016-of-00039.bin",
+    "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00015-of-00039.bin",
+    "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00015-of-00039.bin",
+    "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00015-of-00039.bin",
+    "model.layers.12.self_attn.rotary_emb.inv_freq": "pytorch_model-00015-of-00039.bin",
+    "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00015-of-00039.bin",
+    "model.layers.13.input_layernorm.weight": "pytorch_model-00017-of-00039.bin",
+    "model.layers.13.mlp.down_proj.weight": "pytorch_model-00017-of-00039.bin",
+    "model.layers.13.mlp.gate_proj.weight": "pytorch_model-00017-of-00039.bin",
+    "model.layers.13.mlp.up_proj.weight": "pytorch_model-00017-of-00039.bin",
+    "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00017-of-00039.bin",
+    "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00016-of-00039.bin",
+    "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00017-of-00039.bin",
+    "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00016-of-00039.bin",
+    "model.layers.13.self_attn.rotary_emb.inv_freq": "pytorch_model-00017-of-00039.bin",
+    "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00016-of-00039.bin",
+    "model.layers.14.input_layernorm.weight": "pytorch_model-00018-of-00039.bin",
+    "model.layers.14.mlp.down_proj.weight": "pytorch_model-00018-of-00039.bin",
+    "model.layers.14.mlp.gate_proj.weight": "pytorch_model-00018-of-00039.bin",
+    "model.layers.14.mlp.up_proj.weight": "pytorch_model-00018-of-00039.bin",
+    "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00018-of-00039.bin",
+    "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00017-of-00039.bin",
+    "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00018-of-00039.bin",
+    "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00017-of-00039.bin",
+    "model.layers.14.self_attn.rotary_emb.inv_freq": "pytorch_model-00018-of-00039.bin",
+    "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00018-of-00039.bin",
+    "model.layers.15.input_layernorm.weight": "pytorch_model-00019-of-00039.bin",
+    "model.layers.15.mlp.down_proj.weight": "pytorch_model-00019-of-00039.bin",
+    "model.layers.15.mlp.gate_proj.weight": "pytorch_model-00019-of-00039.bin",
+    "model.layers.15.mlp.up_proj.weight": "pytorch_model-00019-of-00039.bin",
+    "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00019-of-00039.bin",
+    "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00019-of-00039.bin",
+    "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00019-of-00039.bin",
+    "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00018-of-00039.bin",
+    "model.layers.15.self_attn.rotary_emb.inv_freq": "pytorch_model-00019-of-00039.bin",
+    "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00019-of-00039.bin",
+    "model.layers.16.input_layernorm.weight": "pytorch_model-00021-of-00039.bin",
+    "model.layers.16.mlp.down_proj.weight": "pytorch_model-00020-of-00039.bin",
+    "model.layers.16.mlp.gate_proj.weight": "pytorch_model-00020-of-00039.bin",
+    "model.layers.16.mlp.up_proj.weight": "pytorch_model-00021-of-00039.bin",
+    "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00021-of-00039.bin",
+    "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00020-of-00039.bin",
+    "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00020-of-00039.bin",
+    "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00020-of-00039.bin",
+    "model.layers.16.self_attn.rotary_emb.inv_freq": "pytorch_model-00020-of-00039.bin",
+    "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00020-of-00039.bin",
+    "model.layers.17.input_layernorm.weight": "pytorch_model-00022-of-00039.bin",
+    "model.layers.17.mlp.down_proj.weight": "pytorch_model-00022-of-00039.bin",
+    "model.layers.17.mlp.gate_proj.weight": "pytorch_model-00021-of-00039.bin",
+    "model.layers.17.mlp.up_proj.weight": "pytorch_model-00022-of-00039.bin",
+    "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00022-of-00039.bin",
+    "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00021-of-00039.bin",
+    "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00021-of-00039.bin",
+    "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00021-of-00039.bin",
+    "model.layers.17.self_attn.rotary_emb.inv_freq": "pytorch_model-00021-of-00039.bin",
+    "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00021-of-00039.bin",
+    "model.layers.18.input_layernorm.weight": "pytorch_model-00023-of-00039.bin",
+    "model.layers.18.mlp.down_proj.weight": "pytorch_model-00023-of-00039.bin",
+    "model.layers.18.mlp.gate_proj.weight": "pytorch_model-00023-of-00039.bin",
+    "model.layers.18.mlp.up_proj.weight": "pytorch_model-00023-of-00039.bin",
+    "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00023-of-00039.bin",
+    "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00022-of-00039.bin",
+    "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00022-of-00039.bin",
+    "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00022-of-00039.bin",
+    "model.layers.18.self_attn.rotary_emb.inv_freq": "pytorch_model-00022-of-00039.bin",
+    "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00022-of-00039.bin",
+    "model.layers.19.input_layernorm.weight": "pytorch_model-00024-of-00039.bin",
+    "model.layers.19.mlp.down_proj.weight": "pytorch_model-00024-of-00039.bin",
+    "model.layers.19.mlp.gate_proj.weight": "pytorch_model-00024-of-00039.bin",
+    "model.layers.19.mlp.up_proj.weight": "pytorch_model-00024-of-00039.bin",
+    "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00024-of-00039.bin",
+    "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00023-of-00039.bin",
+    "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00024-of-00039.bin",
+    "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00023-of-00039.bin",
+    "model.layers.19.self_attn.rotary_emb.inv_freq": "pytorch_model-00024-of-00039.bin",
+    "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00023-of-00039.bin",
+    "model.layers.2.input_layernorm.weight": "pytorch_model-00004-of-00039.bin",
+    "model.layers.2.mlp.down_proj.weight": "pytorch_model-00004-of-00039.bin",
+    "model.layers.2.mlp.gate_proj.weight": "pytorch_model-00004-of-00039.bin",
+    "model.layers.2.mlp.up_proj.weight": "pytorch_model-00004-of-00039.bin",
+    "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00004-of-00039.bin",
+    "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00003-of-00039.bin",
+    "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00004-of-00039.bin",
+    "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00003-of-00039.bin",
+    "model.layers.2.self_attn.rotary_emb.inv_freq": "pytorch_model-00004-of-00039.bin",
+    "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00004-of-00039.bin",
+    "model.layers.20.input_layernorm.weight": "pytorch_model-00025-of-00039.bin",
+    "model.layers.20.mlp.down_proj.weight": "pytorch_model-00025-of-00039.bin",
+    "model.layers.20.mlp.gate_proj.weight": "pytorch_model-00025-of-00039.bin",
+    "model.layers.20.mlp.up_proj.weight": "pytorch_model-00025-of-00039.bin",
+    "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00025-of-00039.bin",
+    "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00024-of-00039.bin",
+    "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00025-of-00039.bin",
+    "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00024-of-00039.bin",
+    "model.layers.20.self_attn.rotary_emb.inv_freq": "pytorch_model-00025-of-00039.bin",
+    "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00025-of-00039.bin",
+    "model.layers.21.input_layernorm.weight": "pytorch_model-00026-of-00039.bin",
+    "model.layers.21.mlp.down_proj.weight": "pytorch_model-00026-of-00039.bin",
+    "model.layers.21.mlp.gate_proj.weight": "pytorch_model-00026-of-00039.bin",
+    "model.layers.21.mlp.up_proj.weight": "pytorch_model-00026-of-00039.bin",
+    "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00026-of-00039.bin",
+    "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00026-of-00039.bin",
+    "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00026-of-00039.bin",
+    "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00025-of-00039.bin",
+    "model.layers.21.self_attn.rotary_emb.inv_freq": "pytorch_model-00026-of-00039.bin",
+    "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00026-of-00039.bin",
+    "model.layers.22.input_layernorm.weight": "pytorch_model-00028-of-00039.bin",
+    "model.layers.22.mlp.down_proj.weight": "pytorch_model-00027-of-00039.bin",
+    "model.layers.22.mlp.gate_proj.weight": "pytorch_model-00027-of-00039.bin",
+    "model.layers.22.mlp.up_proj.weight": "pytorch_model-00028-of-00039.bin",
+    "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00028-of-00039.bin",
+    "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00027-of-00039.bin",
+    "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00027-of-00039.bin",
+    "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00027-of-00039.bin",
+    "model.layers.22.self_attn.rotary_emb.inv_freq": "pytorch_model-00027-of-00039.bin",
+    "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00027-of-00039.bin",
+    "model.layers.23.input_layernorm.weight": "pytorch_model-00029-of-00039.bin",
+    "model.layers.23.mlp.down_proj.weight": "pytorch_model-00029-of-00039.bin",
+    "model.layers.23.mlp.gate_proj.weight": "pytorch_model-00028-of-00039.bin",
+    "model.layers.23.mlp.up_proj.weight": "pytorch_model-00029-of-00039.bin",
+    "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00029-of-00039.bin",
+    "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00028-of-00039.bin",
+    "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00028-of-00039.bin",
+    "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00028-of-00039.bin",
+    "model.layers.23.self_attn.rotary_emb.inv_freq": "pytorch_model-00028-of-00039.bin",
+    "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00028-of-00039.bin",
+    "model.layers.24.input_layernorm.weight": "pytorch_model-00030-of-00039.bin",
+    "model.layers.24.mlp.down_proj.weight": "pytorch_model-00030-of-00039.bin",
+    "model.layers.24.mlp.gate_proj.weight": "pytorch_model-00030-of-00039.bin",
+    "model.layers.24.mlp.up_proj.weight": "pytorch_model-00030-of-00039.bin",
+    "model.layers.24.post_attention_layernorm.weight": "pytorch_model-00030-of-00039.bin",
+    "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00029-of-00039.bin",
+    "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00029-of-00039.bin",
+    "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00029-of-00039.bin",
+    "model.layers.24.self_attn.rotary_emb.inv_freq": "pytorch_model-00029-of-00039.bin",
+    "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00029-of-00039.bin",
+    "model.layers.25.input_layernorm.weight": "pytorch_model-00031-of-00039.bin",
+    "model.layers.25.mlp.down_proj.weight": "pytorch_model-00031-of-00039.bin",
+    "model.layers.25.mlp.gate_proj.weight": "pytorch_model-00031-of-00039.bin",
+    "model.layers.25.mlp.up_proj.weight": "pytorch_model-00031-of-00039.bin",
+    "model.layers.25.post_attention_layernorm.weight": "pytorch_model-00031-of-00039.bin",
+    "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00030-of-00039.bin",
+    "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00031-of-00039.bin",
+    "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00030-of-00039.bin",
+    "model.layers.25.self_attn.rotary_emb.inv_freq": "pytorch_model-00031-of-00039.bin",
+    "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00030-of-00039.bin",
+    "model.layers.26.input_layernorm.weight": "pytorch_model-00032-of-00039.bin",
+    "model.layers.26.mlp.down_proj.weight": "pytorch_model-00032-of-00039.bin",
+    "model.layers.26.mlp.gate_proj.weight": "pytorch_model-00032-of-00039.bin",
+    "model.layers.26.mlp.up_proj.weight": "pytorch_model-00032-of-00039.bin",
+    "model.layers.26.post_attention_layernorm.weight": "pytorch_model-00032-of-00039.bin",
+    "model.layers.26.self_attn.k_proj.weight": "pytorch_model-00031-of-00039.bin",
+    "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00032-of-00039.bin",
+    "model.layers.26.self_attn.q_proj.weight": "pytorch_model-00031-of-00039.bin",
+    "model.layers.26.self_attn.rotary_emb.inv_freq": "pytorch_model-00032-of-00039.bin",
+    "model.layers.26.self_attn.v_proj.weight": "pytorch_model-00032-of-00039.bin",
+    "model.layers.27.input_layernorm.weight": "pytorch_model-00033-of-00039.bin",
+    "model.layers.27.mlp.down_proj.weight": "pytorch_model-00033-of-00039.bin",
+    "model.layers.27.mlp.gate_proj.weight": "pytorch_model-00033-of-00039.bin",
+    "model.layers.27.mlp.up_proj.weight": "pytorch_model-00033-of-00039.bin",
+    "model.layers.27.post_attention_layernorm.weight": "pytorch_model-00033-of-00039.bin",
+    "model.layers.27.self_attn.k_proj.weight": "pytorch_model-00033-of-00039.bin",
+    "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00033-of-00039.bin",
+    "model.layers.27.self_attn.q_proj.weight": "pytorch_model-00032-of-00039.bin",
+    "model.layers.27.self_attn.rotary_emb.inv_freq": "pytorch_model-00033-of-00039.bin",
+    "model.layers.27.self_attn.v_proj.weight": "pytorch_model-00033-of-00039.bin",
+    "model.layers.28.input_layernorm.weight": "pytorch_model-00035-of-00039.bin",
+    "model.layers.28.mlp.down_proj.weight": "pytorch_model-00034-of-00039.bin",
+    "model.layers.28.mlp.gate_proj.weight": "pytorch_model-00034-of-00039.bin",
+    "model.layers.28.mlp.up_proj.weight": "pytorch_model-00035-of-00039.bin",
+    "model.layers.28.post_attention_layernorm.weight": "pytorch_model-00035-of-00039.bin",
+    "model.layers.28.self_attn.k_proj.weight": "pytorch_model-00034-of-00039.bin",
+    "model.layers.28.self_attn.o_proj.weight": "pytorch_model-00034-of-00039.bin",
+    "model.layers.28.self_attn.q_proj.weight": "pytorch_model-00034-of-00039.bin",
+    "model.layers.28.self_attn.rotary_emb.inv_freq": "pytorch_model-00034-of-00039.bin",
+    "model.layers.28.self_attn.v_proj.weight": "pytorch_model-00034-of-00039.bin",
+    "model.layers.29.input_layernorm.weight": "pytorch_model-00036-of-00039.bin",
+    "model.layers.29.mlp.down_proj.weight": "pytorch_model-00036-of-00039.bin",
+    "model.layers.29.mlp.gate_proj.weight": "pytorch_model-00035-of-00039.bin",
+    "model.layers.29.mlp.up_proj.weight": "pytorch_model-00036-of-00039.bin",
+    "model.layers.29.post_attention_layernorm.weight": "pytorch_model-00036-of-00039.bin",
+    "model.layers.29.self_attn.k_proj.weight": "pytorch_model-00035-of-00039.bin",
+    "model.layers.29.self_attn.o_proj.weight": "pytorch_model-00035-of-00039.bin",
+    "model.layers.29.self_attn.q_proj.weight": "pytorch_model-00035-of-00039.bin",
+    "model.layers.29.self_attn.rotary_emb.inv_freq": "pytorch_model-00035-of-00039.bin",
+    "model.layers.29.self_attn.v_proj.weight": "pytorch_model-00035-of-00039.bin",
+    "model.layers.3.input_layernorm.weight": "pytorch_model-00005-of-00039.bin",
+    "model.layers.3.mlp.down_proj.weight": "pytorch_model-00005-of-00039.bin",
+    "model.layers.3.mlp.gate_proj.weight": "pytorch_model-00005-of-00039.bin",
+    "model.layers.3.mlp.up_proj.weight": "pytorch_model-00005-of-00039.bin",
+    "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00005-of-00039.bin",
+    "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00005-of-00039.bin",
+    "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00005-of-00039.bin",
+    "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00004-of-00039.bin",
+    "model.layers.3.self_attn.rotary_emb.inv_freq": "pytorch_model-00005-of-00039.bin",
+    "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00005-of-00039.bin",
+    "model.layers.30.input_layernorm.weight": "pytorch_model-00037-of-00039.bin",
+    "model.layers.30.mlp.down_proj.weight": "pytorch_model-00037-of-00039.bin",
+    "model.layers.30.mlp.gate_proj.weight": "pytorch_model-00037-of-00039.bin",
+    "model.layers.30.mlp.up_proj.weight": "pytorch_model-00037-of-00039.bin",
+    "model.layers.30.post_attention_layernorm.weight": "pytorch_model-00037-of-00039.bin",
+    "model.layers.30.self_attn.k_proj.weight": "pytorch_model-00036-of-00039.bin",
+    "model.layers.30.self_attn.o_proj.weight": "pytorch_model-00036-of-00039.bin",
+    "model.layers.30.self_attn.q_proj.weight": "pytorch_model-00036-of-00039.bin",
+    "model.layers.30.self_attn.rotary_emb.inv_freq": "pytorch_model-00036-of-00039.bin",
+    "model.layers.30.self_attn.v_proj.weight": "pytorch_model-00036-of-00039.bin",
+    "model.layers.31.input_layernorm.weight": "pytorch_model-00038-of-00039.bin",
+    "model.layers.31.mlp.down_proj.weight": "pytorch_model-00038-of-00039.bin",
+    "model.layers.31.mlp.gate_proj.weight": "pytorch_model-00038-of-00039.bin",
+    "model.layers.31.mlp.up_proj.weight": "pytorch_model-00038-of-00039.bin",
+    "model.layers.31.post_attention_layernorm.weight": "pytorch_model-00038-of-00039.bin",
+    "model.layers.31.self_attn.k_proj.weight": "pytorch_model-00037-of-00039.bin",
+    "model.layers.31.self_attn.o_proj.weight": "pytorch_model-00038-of-00039.bin",
+    "model.layers.31.self_attn.q_proj.weight": "pytorch_model-00037-of-00039.bin",
+    "model.layers.31.self_attn.rotary_emb.inv_freq": "pytorch_model-00038-of-00039.bin",
+    "model.layers.31.self_attn.v_proj.weight": "pytorch_model-00037-of-00039.bin",
+    "model.layers.4.input_layernorm.weight": "pytorch_model-00007-of-00039.bin",
+    "model.layers.4.mlp.down_proj.weight": "pytorch_model-00006-of-00039.bin",
+    "model.layers.4.mlp.gate_proj.weight": "pytorch_model-00006-of-00039.bin",
+    "model.layers.4.mlp.up_proj.weight": "pytorch_model-00007-of-00039.bin",
+    "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00007-of-00039.bin",
+    "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00006-of-00039.bin",
+    "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00006-of-00039.bin",
+    "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00006-of-00039.bin",
+    "model.layers.4.self_attn.rotary_emb.inv_freq": "pytorch_model-00006-of-00039.bin",
+    "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00006-of-00039.bin",
+    "model.layers.5.input_layernorm.weight": "pytorch_model-00008-of-00039.bin",
+    "model.layers.5.mlp.down_proj.weight": "pytorch_model-00008-of-00039.bin",
+    "model.layers.5.mlp.gate_proj.weight": "pytorch_model-00007-of-00039.bin",
+    "model.layers.5.mlp.up_proj.weight": "pytorch_model-00008-of-00039.bin",
+    "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00008-of-00039.bin",
+    "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00007-of-00039.bin",
+    "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00007-of-00039.bin",
+    "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00007-of-00039.bin",
+    "model.layers.5.self_attn.rotary_emb.inv_freq": "pytorch_model-00007-of-00039.bin",
+    "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00007-of-00039.bin",
+    "model.layers.6.input_layernorm.weight": "pytorch_model-00009-of-00039.bin",
+    "model.layers.6.mlp.down_proj.weight": "pytorch_model-00009-of-00039.bin",
+    "model.layers.6.mlp.gate_proj.weight": "pytorch_model-00009-of-00039.bin",
+    "model.layers.6.mlp.up_proj.weight": "pytorch_model-00009-of-00039.bin",
+    "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00009-of-00039.bin",
+    "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00008-of-00039.bin",
+    "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00008-of-00039.bin",
+    "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00008-of-00039.bin",
+    "model.layers.6.self_attn.rotary_emb.inv_freq": "pytorch_model-00008-of-00039.bin",
+    "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00008-of-00039.bin",
+    "model.layers.7.input_layernorm.weight": "pytorch_model-00010-of-00039.bin",
+    "model.layers.7.mlp.down_proj.weight": "pytorch_model-00010-of-00039.bin",
+    "model.layers.7.mlp.gate_proj.weight": "pytorch_model-00010-of-00039.bin",
+    "model.layers.7.mlp.up_proj.weight": "pytorch_model-00010-of-00039.bin",
+    "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00010-of-00039.bin",
+    "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00009-of-00039.bin",
+    "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00010-of-00039.bin",
+    "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00009-of-00039.bin",
+    "model.layers.7.self_attn.rotary_emb.inv_freq": "pytorch_model-00010-of-00039.bin",
+    "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00009-of-00039.bin",
+    "model.layers.8.input_layernorm.weight": "pytorch_model-00011-of-00039.bin",
+    "model.layers.8.mlp.down_proj.weight": "pytorch_model-00011-of-00039.bin",
+    "model.layers.8.mlp.gate_proj.weight": "pytorch_model-00011-of-00039.bin",
+    "model.layers.8.mlp.up_proj.weight": "pytorch_model-00011-of-00039.bin",
+    "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00011-of-00039.bin",
+    "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00010-of-00039.bin",
+    "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00011-of-00039.bin",
+    "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00010-of-00039.bin",
+    "model.layers.8.self_attn.rotary_emb.inv_freq": "pytorch_model-00011-of-00039.bin",
+    "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00011-of-00039.bin",
+    "model.layers.9.input_layernorm.weight": "pytorch_model-00012-of-00039.bin",
+    "model.layers.9.mlp.down_proj.weight": "pytorch_model-00012-of-00039.bin",
+    "model.layers.9.mlp.gate_proj.weight": "pytorch_model-00012-of-00039.bin",
+    "model.layers.9.mlp.up_proj.weight": "pytorch_model-00012-of-00039.bin",
+    "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00012-of-00039.bin",
+    "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00012-of-00039.bin",
+    "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00012-of-00039.bin",
+    "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00011-of-00039.bin",
+    "model.layers.9.self_attn.rotary_emb.inv_freq": "pytorch_model-00012-of-00039.bin",
+    "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00012-of-00039.bin",
+    "model.norm.weight": "pytorch_model-00038-of-00039.bin"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "bos_token": {
+    "content": "",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

tokenizer_config.json

@@ -0,0 +1,34 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "bos_token": {
+    "__type": "AddedToken",
+    "content": "",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "clean_up_tokenization_spaces": false,
+  "eos_token": {
+    "__type": "AddedToken",
+    "content": "",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": null,
+  "sp_model_kwargs": {},
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": {
+    "__type": "AddedToken",
+    "content": "",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "use_fast": false
+}