Hugging Face's logo

fbaldassarri
/
EleutherAI_pythia-12b-deduped-autoround-int4-gs128-sym

Text Generation
Safetensors
PyTorch
English
gpt_neox
causal-lm
pythia
autoround
intel
woq
intel-autoround
4-bit precision
intel/auto-round
Model card Files Community
fbaldassarri commited on
Commit
e6c2d15
•
1 Parent(s): 67170e5

Initial Upload

Browse files
Files changed (10) hide show
  1. README.md +87 -3
  2. config.json +56 -0
  3. generation_config.json +6 -0
  4. model-00001-of-00002.safetensors +3 -0
  5. model-00002-of-00002.safetensors +3 -0
  6. model.safetensors.index.json +875 -0
  7. quantization_config.json +24 -0
  8. special_tokens_map.json +23 -0
  9. tokenizer.json +0 -0
  10. tokenizer_config.json +214 -0
README.md CHANGED
@@ -1,3 +1,87 @@
1
- ---
2
- license: apache-2.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ tags:
5
+ - pytorch
6
+ - causal-lm
7
+ - pythia
8
+ - autoround
9
+ - intel
10
+ - woq
11
+ - intel-autoround
12
+ license: apache-2.0
13
+ model_name: Pythia 12b deduped
14
+ base_model: EleutherAI/pythia-12b-deduped
15
+ inference: false
16
+ model_creator: EleutherAI
17
+ datasets:
18
+ - EleutherAI/pile
19
+ pipeline_tag: text-generation
20
+ prompt_template: '{prompt}
21
+ '
22
+ quantized_by: fbaldassarri
23
+ ---
24
+
25
+
26
+
27
+ ## Model Information
28
+
29
+ Quantized version of [EleutherAI/pythia-12b-deduped](https://huggingface.co/EleutherAI/pythia-12b-deduped) using torch.float32 for quantization tuning.
30
+ - 4 bits (INT4)
31
+ - group size = 128
32
+ - Symmetrical Quantization
33
+ - Method AutoRound (WOQ)
34
+
35
+ Fast and low memory, 2-3X speedup (slight accuracy drop at W4G128)
36
+
37
+ Quantization framework: [Intel AutoRound](https://github.com/intel/auto-round)
38
+
39
+ Note: this INT4 version of pythia-12b-deduped has been quantized to run inference through CPU.
40
+
41
+ ## Replication Recipe
42
+
43
+ ### Step 1 Install Requirements
44
+
45
+ I suggest to install requirements into a dedicated python-virtualenv or a conda enviroment.
46
+
47
+ ```
48
+ python -m pip install <package> --upgrade
49
+ ```
50
+
51
+ - accelerate==1.0.1
52
+ - auto_gptq==0.7.1
53
+ - neural_compressor==3.1
54
+ - torch==2.3.0+cpu
55
+ - torchaudio==2.5.0+cpu
56
+ - torchvision==0.18.0+cpu
57
+ - transformers==4.45.2
58
+
59
+ ### Step 2 Build Intel AutoRound wheel from sources
60
+
61
+ ```
62
+ python -m pip install git+https://github.com/intel/auto-round.git
63
+ ```
64
+
65
+ ### Step 3 Script for Quantization
66
+
67
+ ```
68
+ from transformers import AutoModelForCausalLM, AutoTokenizer
69
+ model_name = "EleutherAI/pythia-12b-deduped"
70
+ model = AutoModelForCausalLM.from_pretrained(model_name)
71
+ tokenizer = AutoTokenizer.from_pretrained(model_name)
72
+ from auto_round import AutoRound
73
+ bits, group_size, sym = 4, 128, True
74
+ autoround = AutoRound(model, tokenizer, nsamples=128, iters=200, seqlen=512, batch_size=4, bits=bits, group_size=group_size, sym=sym)
75
+ autoround.quantize()
76
+ output_dir = "./AutoRound/EleutherAI_pythia-12b-deduped-autoround-int4-gs128-sym"
77
+ autoround.save_quantized(output_dir, format='auto_round', inplace=True)
78
+ ```
79
+
80
+ ## License
81
+
82
+ [Apache 2.0 License](https://choosealicense.com/licenses/apache-2.0/)
83
+
84
+ ## Disclaimer
85
+
86
+ This quantized model comes with no warranty. It has been developed only for research purposes.
87
+
config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "EleutherAI/pythia-12b-deduped",
3
+ "architectures": [
4
+ "GPTNeoXForCausalLM"
5
+ ],
6
+ "attention_bias": true,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 0,
9
+ "classifier_dropout": 0.1,
10
+ "eos_token_id": 0,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout": 0.0,
13
+ "hidden_size": 5120,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 20480,
16
+ "layer_norm_eps": 1e-05,
17
+ "max_position_embeddings": 2048,
18
+ "model_type": "gpt_neox",
19
+ "num_attention_heads": 40,
20
+ "num_hidden_layers": 36,
21
+ "partial_rotary_factor": 0.25,
22
+ "quantization_config": {
23
+ "amp": false,
24
+ "autoround_version": "0.4.0.dev",
25
+ "backend": "auto_round:exllamav2",
26
+ "bits": 4,
27
+ "data_type": "int",
28
+ "dataset": "NeelNanda/pile-10k",
29
+ "enable_minmax_tuning": true,
30
+ "enable_norm_bias_tuning": false,
31
+ "enable_quanted_input": true,
32
+ "gradient_accumulate_steps": 1,
33
+ "group_size": 128,
34
+ "iters": 200,
35
+ "low_gpu_mem_usage": false,
36
+ "lr": 0.005,
37
+ "minmax_lr": 0.005,
38
+ "nsamples": 128,
39
+ "quant_block_list": null,
40
+ "quant_method": "intel/auto-round",
41
+ "scale_dtype": "torch.float16",
42
+ "seqlen": 512,
43
+ "sym": false,
44
+ "train_bs": 4
45
+ },
46
+ "rope_scaling": null,
47
+ "rope_theta": 10000,
48
+ "rotary_emb_base": 10000,
49
+ "rotary_pct": 0.25,
50
+ "tie_word_embeddings": false,
51
+ "torch_dtype": "float32",
52
+ "transformers_version": "4.45.2",
53
+ "use_cache": true,
54
+ "use_parallel_residual": true,
55
+ "vocab_size": 50688
56
+ }
generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "eos_token_id": 0,
5
+ "transformers_version": "4.45.2"
6
+ }
model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bcbe65aada97eaa299c51f1a36cad8f0167362b2ec353b289ea7b232443433d6
3
+ size 4968189328
model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edd6fbc6864cc80a99f2775ac0b9bb56d96bf8cb07512e70fae51b23733fda2c
3
+ size 3003057544
model.safetensors.index.json ADDED
@@ -0,0 +1,875 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 7971143680
4
+ },
5
+ "weight_map": {
6
+ "embed_out.weight": "model-00002-of-00002.safetensors",
7
+ "gpt_neox.embed_in.weight": "model-00001-of-00002.safetensors",
8
+ "gpt_neox.final_layer_norm.bias": "model-00002-of-00002.safetensors",
9
+ "gpt_neox.final_layer_norm.weight": "model-00002-of-00002.safetensors",
10
+ "gpt_neox.layers.0.attention.dense.bias": "model-00001-of-00002.safetensors",
11
+ "gpt_neox.layers.0.attention.dense.g_idx": "model-00001-of-00002.safetensors",
12
+ "gpt_neox.layers.0.attention.dense.qweight": "model-00001-of-00002.safetensors",
13
+ "gpt_neox.layers.0.attention.dense.qzeros": "model-00001-of-00002.safetensors",
14
+ "gpt_neox.layers.0.attention.dense.scales": "model-00001-of-00002.safetensors",
15
+ "gpt_neox.layers.0.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
16
+ "gpt_neox.layers.0.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
17
+ "gpt_neox.layers.0.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
18
+ "gpt_neox.layers.0.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
19
+ "gpt_neox.layers.0.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
20
+ "gpt_neox.layers.0.input_layernorm.bias": "model-00001-of-00002.safetensors",
21
+ "gpt_neox.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
22
+ "gpt_neox.layers.0.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
23
+ "gpt_neox.layers.0.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
24
+ "gpt_neox.layers.0.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
25
+ "gpt_neox.layers.0.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
26
+ "gpt_neox.layers.0.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
27
+ "gpt_neox.layers.0.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
28
+ "gpt_neox.layers.0.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
29
+ "gpt_neox.layers.0.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
30
+ "gpt_neox.layers.0.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
31
+ "gpt_neox.layers.0.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
32
+ "gpt_neox.layers.0.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
33
+ "gpt_neox.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
34
+ "gpt_neox.layers.1.attention.dense.bias": "model-00001-of-00002.safetensors",
35
+ "gpt_neox.layers.1.attention.dense.g_idx": "model-00001-of-00002.safetensors",
36
+ "gpt_neox.layers.1.attention.dense.qweight": "model-00001-of-00002.safetensors",
37
+ "gpt_neox.layers.1.attention.dense.qzeros": "model-00001-of-00002.safetensors",
38
+ "gpt_neox.layers.1.attention.dense.scales": "model-00001-of-00002.safetensors",
39
+ "gpt_neox.layers.1.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
40
+ "gpt_neox.layers.1.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
41
+ "gpt_neox.layers.1.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
42
+ "gpt_neox.layers.1.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
43
+ "gpt_neox.layers.1.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
44
+ "gpt_neox.layers.1.input_layernorm.bias": "model-00001-of-00002.safetensors",
45
+ "gpt_neox.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
46
+ "gpt_neox.layers.1.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
47
+ "gpt_neox.layers.1.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
48
+ "gpt_neox.layers.1.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
49
+ "gpt_neox.layers.1.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
50
+ "gpt_neox.layers.1.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
51
+ "gpt_neox.layers.1.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
52
+ "gpt_neox.layers.1.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
53
+ "gpt_neox.layers.1.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
54
+ "gpt_neox.layers.1.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
55
+ "gpt_neox.layers.1.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
56
+ "gpt_neox.layers.1.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
57
+ "gpt_neox.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
58
+ "gpt_neox.layers.10.attention.dense.bias": "model-00001-of-00002.safetensors",
59
+ "gpt_neox.layers.10.attention.dense.g_idx": "model-00001-of-00002.safetensors",
60
+ "gpt_neox.layers.10.attention.dense.qweight": "model-00001-of-00002.safetensors",
61
+ "gpt_neox.layers.10.attention.dense.qzeros": "model-00001-of-00002.safetensors",
62
+ "gpt_neox.layers.10.attention.dense.scales": "model-00001-of-00002.safetensors",
63
+ "gpt_neox.layers.10.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
64
+ "gpt_neox.layers.10.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
65
+ "gpt_neox.layers.10.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
66
+ "gpt_neox.layers.10.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
67
+ "gpt_neox.layers.10.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
68
+ "gpt_neox.layers.10.input_layernorm.bias": "model-00001-of-00002.safetensors",
69
+ "gpt_neox.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
70
+ "gpt_neox.layers.10.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
71
+ "gpt_neox.layers.10.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
72
+ "gpt_neox.layers.10.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
73
+ "gpt_neox.layers.10.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
74
+ "gpt_neox.layers.10.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
75
+ "gpt_neox.layers.10.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
76
+ "gpt_neox.layers.10.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
77
+ "gpt_neox.layers.10.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
78
+ "gpt_neox.layers.10.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
79
+ "gpt_neox.layers.10.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
80
+ "gpt_neox.layers.10.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
81
+ "gpt_neox.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
82
+ "gpt_neox.layers.11.attention.dense.bias": "model-00001-of-00002.safetensors",
83
+ "gpt_neox.layers.11.attention.dense.g_idx": "model-00001-of-00002.safetensors",
84
+ "gpt_neox.layers.11.attention.dense.qweight": "model-00001-of-00002.safetensors",
85
+ "gpt_neox.layers.11.attention.dense.qzeros": "model-00001-of-00002.safetensors",
86
+ "gpt_neox.layers.11.attention.dense.scales": "model-00001-of-00002.safetensors",
87
+ "gpt_neox.layers.11.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
88
+ "gpt_neox.layers.11.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
89
+ "gpt_neox.layers.11.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
90
+ "gpt_neox.layers.11.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
91
+ "gpt_neox.layers.11.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
92
+ "gpt_neox.layers.11.input_layernorm.bias": "model-00001-of-00002.safetensors",
93
+ "gpt_neox.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
94
+ "gpt_neox.layers.11.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
95
+ "gpt_neox.layers.11.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
96
+ "gpt_neox.layers.11.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
97
+ "gpt_neox.layers.11.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
98
+ "gpt_neox.layers.11.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
99
+ "gpt_neox.layers.11.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
100
+ "gpt_neox.layers.11.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
101
+ "gpt_neox.layers.11.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
102
+ "gpt_neox.layers.11.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
103
+ "gpt_neox.layers.11.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
104
+ "gpt_neox.layers.11.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
105
+ "gpt_neox.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
106
+ "gpt_neox.layers.12.attention.dense.bias": "model-00001-of-00002.safetensors",
107
+ "gpt_neox.layers.12.attention.dense.g_idx": "model-00001-of-00002.safetensors",
108
+ "gpt_neox.layers.12.attention.dense.qweight": "model-00001-of-00002.safetensors",
109
+ "gpt_neox.layers.12.attention.dense.qzeros": "model-00001-of-00002.safetensors",
110
+ "gpt_neox.layers.12.attention.dense.scales": "model-00001-of-00002.safetensors",
111
+ "gpt_neox.layers.12.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
112
+ "gpt_neox.layers.12.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
113
+ "gpt_neox.layers.12.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
114
+ "gpt_neox.layers.12.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
115
+ "gpt_neox.layers.12.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
116
+ "gpt_neox.layers.12.input_layernorm.bias": "model-00001-of-00002.safetensors",
117
+ "gpt_neox.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
118
+ "gpt_neox.layers.12.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
119
+ "gpt_neox.layers.12.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
120
+ "gpt_neox.layers.12.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
121
+ "gpt_neox.layers.12.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
122
+ "gpt_neox.layers.12.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
123
+ "gpt_neox.layers.12.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
124
+ "gpt_neox.layers.12.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
125
+ "gpt_neox.layers.12.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
126
+ "gpt_neox.layers.12.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
127
+ "gpt_neox.layers.12.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
128
+ "gpt_neox.layers.12.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
129
+ "gpt_neox.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
130
+ "gpt_neox.layers.13.attention.dense.bias": "model-00001-of-00002.safetensors",
131
+ "gpt_neox.layers.13.attention.dense.g_idx": "model-00001-of-00002.safetensors",
132
+ "gpt_neox.layers.13.attention.dense.qweight": "model-00001-of-00002.safetensors",
133
+ "gpt_neox.layers.13.attention.dense.qzeros": "model-00001-of-00002.safetensors",
134
+ "gpt_neox.layers.13.attention.dense.scales": "model-00001-of-00002.safetensors",
135
+ "gpt_neox.layers.13.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
136
+ "gpt_neox.layers.13.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
137
+ "gpt_neox.layers.13.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
138
+ "gpt_neox.layers.13.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
139
+ "gpt_neox.layers.13.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
140
+ "gpt_neox.layers.13.input_layernorm.bias": "model-00001-of-00002.safetensors",
141
+ "gpt_neox.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
142
+ "gpt_neox.layers.13.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
143
+ "gpt_neox.layers.13.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
144
+ "gpt_neox.layers.13.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
145
+ "gpt_neox.layers.13.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
146
+ "gpt_neox.layers.13.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
147
+ "gpt_neox.layers.13.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
148
+ "gpt_neox.layers.13.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
149
+ "gpt_neox.layers.13.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
150
+ "gpt_neox.layers.13.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
151
+ "gpt_neox.layers.13.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
152
+ "gpt_neox.layers.13.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
153
+ "gpt_neox.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
154
+ "gpt_neox.layers.14.attention.dense.bias": "model-00001-of-00002.safetensors",
155
+ "gpt_neox.layers.14.attention.dense.g_idx": "model-00001-of-00002.safetensors",
156
+ "gpt_neox.layers.14.attention.dense.qweight": "model-00001-of-00002.safetensors",
157
+ "gpt_neox.layers.14.attention.dense.qzeros": "model-00001-of-00002.safetensors",
158
+ "gpt_neox.layers.14.attention.dense.scales": "model-00001-of-00002.safetensors",
159
+ "gpt_neox.layers.14.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
160
+ "gpt_neox.layers.14.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
161
+ "gpt_neox.layers.14.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
162
+ "gpt_neox.layers.14.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
163
+ "gpt_neox.layers.14.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
164
+ "gpt_neox.layers.14.input_layernorm.bias": "model-00001-of-00002.safetensors",
165
+ "gpt_neox.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
166
+ "gpt_neox.layers.14.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
167
+ "gpt_neox.layers.14.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
168
+ "gpt_neox.layers.14.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
169
+ "gpt_neox.layers.14.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
170
+ "gpt_neox.layers.14.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
171
+ "gpt_neox.layers.14.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
172
+ "gpt_neox.layers.14.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
173
+ "gpt_neox.layers.14.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
174
+ "gpt_neox.layers.14.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
175
+ "gpt_neox.layers.14.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
176
+ "gpt_neox.layers.14.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
177
+ "gpt_neox.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
178
+ "gpt_neox.layers.15.attention.dense.bias": "model-00001-of-00002.safetensors",
179
+ "gpt_neox.layers.15.attention.dense.g_idx": "model-00001-of-00002.safetensors",
180
+ "gpt_neox.layers.15.attention.dense.qweight": "model-00001-of-00002.safetensors",
181
+ "gpt_neox.layers.15.attention.dense.qzeros": "model-00001-of-00002.safetensors",
182
+ "gpt_neox.layers.15.attention.dense.scales": "model-00001-of-00002.safetensors",
183
+ "gpt_neox.layers.15.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
184
+ "gpt_neox.layers.15.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
185
+ "gpt_neox.layers.15.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
186
+ "gpt_neox.layers.15.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
187
+ "gpt_neox.layers.15.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
188
+ "gpt_neox.layers.15.input_layernorm.bias": "model-00001-of-00002.safetensors",
189
+ "gpt_neox.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
190
+ "gpt_neox.layers.15.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
191
+ "gpt_neox.layers.15.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
192
+ "gpt_neox.layers.15.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
193
+ "gpt_neox.layers.15.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
194
+ "gpt_neox.layers.15.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
195
+ "gpt_neox.layers.15.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
196
+ "gpt_neox.layers.15.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
197
+ "gpt_neox.layers.15.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
198
+ "gpt_neox.layers.15.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
199
+ "gpt_neox.layers.15.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
200
+ "gpt_neox.layers.15.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
201
+ "gpt_neox.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
202
+ "gpt_neox.layers.16.attention.dense.bias": "model-00001-of-00002.safetensors",
203
+ "gpt_neox.layers.16.attention.dense.g_idx": "model-00001-of-00002.safetensors",
204
+ "gpt_neox.layers.16.attention.dense.qweight": "model-00001-of-00002.safetensors",
205
+ "gpt_neox.layers.16.attention.dense.qzeros": "model-00001-of-00002.safetensors",
206
+ "gpt_neox.layers.16.attention.dense.scales": "model-00001-of-00002.safetensors",
207
+ "gpt_neox.layers.16.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
208
+ "gpt_neox.layers.16.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
209
+ "gpt_neox.layers.16.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
210
+ "gpt_neox.layers.16.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
211
+ "gpt_neox.layers.16.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
212
+ "gpt_neox.layers.16.input_layernorm.bias": "model-00001-of-00002.safetensors",
213
+ "gpt_neox.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
214
+ "gpt_neox.layers.16.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
215
+ "gpt_neox.layers.16.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
216
+ "gpt_neox.layers.16.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
217
+ "gpt_neox.layers.16.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
218
+ "gpt_neox.layers.16.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
219
+ "gpt_neox.layers.16.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
220
+ "gpt_neox.layers.16.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
221
+ "gpt_neox.layers.16.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
222
+ "gpt_neox.layers.16.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
223
+ "gpt_neox.layers.16.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
224
+ "gpt_neox.layers.16.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
225
+ "gpt_neox.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
226
+ "gpt_neox.layers.17.attention.dense.bias": "model-00001-of-00002.safetensors",
227
+ "gpt_neox.layers.17.attention.dense.g_idx": "model-00001-of-00002.safetensors",
228
+ "gpt_neox.layers.17.attention.dense.qweight": "model-00001-of-00002.safetensors",
229
+ "gpt_neox.layers.17.attention.dense.qzeros": "model-00001-of-00002.safetensors",
230
+ "gpt_neox.layers.17.attention.dense.scales": "model-00001-of-00002.safetensors",
231
+ "gpt_neox.layers.17.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
232
+ "gpt_neox.layers.17.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
233
+ "gpt_neox.layers.17.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
234
+ "gpt_neox.layers.17.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
235
+ "gpt_neox.layers.17.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
236
+ "gpt_neox.layers.17.input_layernorm.bias": "model-00001-of-00002.safetensors",
237
+ "gpt_neox.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
238
+ "gpt_neox.layers.17.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
239
+ "gpt_neox.layers.17.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
240
+ "gpt_neox.layers.17.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
241
+ "gpt_neox.layers.17.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
242
+ "gpt_neox.layers.17.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
243
+ "gpt_neox.layers.17.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
244
+ "gpt_neox.layers.17.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
245
+ "gpt_neox.layers.17.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
246
+ "gpt_neox.layers.17.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
247
+ "gpt_neox.layers.17.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
248
+ "gpt_neox.layers.17.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
249
+ "gpt_neox.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
250
+ "gpt_neox.layers.18.attention.dense.bias": "model-00001-of-00002.safetensors",
251
+ "gpt_neox.layers.18.attention.dense.g_idx": "model-00001-of-00002.safetensors",
252
+ "gpt_neox.layers.18.attention.dense.qweight": "model-00001-of-00002.safetensors",
253
+ "gpt_neox.layers.18.attention.dense.qzeros": "model-00001-of-00002.safetensors",
254
+ "gpt_neox.layers.18.attention.dense.scales": "model-00001-of-00002.safetensors",
255
+ "gpt_neox.layers.18.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
256
+ "gpt_neox.layers.18.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
257
+ "gpt_neox.layers.18.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
258
+ "gpt_neox.layers.18.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
259
+ "gpt_neox.layers.18.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
260
+ "gpt_neox.layers.18.input_layernorm.bias": "model-00001-of-00002.safetensors",
261
+ "gpt_neox.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
262
+ "gpt_neox.layers.18.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
263
+ "gpt_neox.layers.18.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
264
+ "gpt_neox.layers.18.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
265
+ "gpt_neox.layers.18.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
266
+ "gpt_neox.layers.18.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
267
+ "gpt_neox.layers.18.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
268
+ "gpt_neox.layers.18.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
269
+ "gpt_neox.layers.18.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
270
+ "gpt_neox.layers.18.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
271
+ "gpt_neox.layers.18.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
272
+ "gpt_neox.layers.18.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
273
+ "gpt_neox.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
274
+ "gpt_neox.layers.19.attention.dense.bias": "model-00001-of-00002.safetensors",
275
+ "gpt_neox.layers.19.attention.dense.g_idx": "model-00001-of-00002.safetensors",
276
+ "gpt_neox.layers.19.attention.dense.qweight": "model-00001-of-00002.safetensors",
277
+ "gpt_neox.layers.19.attention.dense.qzeros": "model-00001-of-00002.safetensors",
278
+ "gpt_neox.layers.19.attention.dense.scales": "model-00001-of-00002.safetensors",
279
+ "gpt_neox.layers.19.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
280
+ "gpt_neox.layers.19.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
281
+ "gpt_neox.layers.19.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
282
+ "gpt_neox.layers.19.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
283
+ "gpt_neox.layers.19.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
284
+ "gpt_neox.layers.19.input_layernorm.bias": "model-00001-of-00002.safetensors",
285
+ "gpt_neox.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
286
+ "gpt_neox.layers.19.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
287
+ "gpt_neox.layers.19.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
288
+ "gpt_neox.layers.19.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
289
+ "gpt_neox.layers.19.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
290
+ "gpt_neox.layers.19.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
291
+ "gpt_neox.layers.19.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
292
+ "gpt_neox.layers.19.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
293
+ "gpt_neox.layers.19.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
294
+ "gpt_neox.layers.19.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
295
+ "gpt_neox.layers.19.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
296
+ "gpt_neox.layers.19.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
297
+ "gpt_neox.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
298
+ "gpt_neox.layers.2.attention.dense.bias": "model-00001-of-00002.safetensors",
299
+ "gpt_neox.layers.2.attention.dense.g_idx": "model-00001-of-00002.safetensors",
300
+ "gpt_neox.layers.2.attention.dense.qweight": "model-00001-of-00002.safetensors",
301
+ "gpt_neox.layers.2.attention.dense.qzeros": "model-00001-of-00002.safetensors",
302
+ "gpt_neox.layers.2.attention.dense.scales": "model-00001-of-00002.safetensors",
303
+ "gpt_neox.layers.2.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
304
+ "gpt_neox.layers.2.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
305
+ "gpt_neox.layers.2.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
306
+ "gpt_neox.layers.2.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
307
+ "gpt_neox.layers.2.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
308
+ "gpt_neox.layers.2.input_layernorm.bias": "model-00001-of-00002.safetensors",
309
+ "gpt_neox.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
310
+ "gpt_neox.layers.2.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
311
+ "gpt_neox.layers.2.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
312
+ "gpt_neox.layers.2.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
313
+ "gpt_neox.layers.2.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
314
+ "gpt_neox.layers.2.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
315
+ "gpt_neox.layers.2.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
316
+ "gpt_neox.layers.2.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
317
+ "gpt_neox.layers.2.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
318
+ "gpt_neox.layers.2.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
319
+ "gpt_neox.layers.2.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
320
+ "gpt_neox.layers.2.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
321
+ "gpt_neox.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
322
+ "gpt_neox.layers.20.attention.dense.bias": "model-00001-of-00002.safetensors",
323
+ "gpt_neox.layers.20.attention.dense.g_idx": "model-00001-of-00002.safetensors",
324
+ "gpt_neox.layers.20.attention.dense.qweight": "model-00001-of-00002.safetensors",
325
+ "gpt_neox.layers.20.attention.dense.qzeros": "model-00001-of-00002.safetensors",
326
+ "gpt_neox.layers.20.attention.dense.scales": "model-00001-of-00002.safetensors",
327
+ "gpt_neox.layers.20.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
328
+ "gpt_neox.layers.20.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
329
+ "gpt_neox.layers.20.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
330
+ "gpt_neox.layers.20.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
331
+ "gpt_neox.layers.20.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
332
+ "gpt_neox.layers.20.input_layernorm.bias": "model-00001-of-00002.safetensors",
333
+ "gpt_neox.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
334
+ "gpt_neox.layers.20.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
335
+ "gpt_neox.layers.20.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
336
+ "gpt_neox.layers.20.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
337
+ "gpt_neox.layers.20.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
338
+ "gpt_neox.layers.20.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
339
+ "gpt_neox.layers.20.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
340
+ "gpt_neox.layers.20.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
341
+ "gpt_neox.layers.20.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
342
+ "gpt_neox.layers.20.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
343
+ "gpt_neox.layers.20.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
344
+ "gpt_neox.layers.20.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
345
+ "gpt_neox.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
346
+ "gpt_neox.layers.21.attention.dense.bias": "model-00001-of-00002.safetensors",
347
+ "gpt_neox.layers.21.attention.dense.g_idx": "model-00001-of-00002.safetensors",
348
+ "gpt_neox.layers.21.attention.dense.qweight": "model-00001-of-00002.safetensors",
349
+ "gpt_neox.layers.21.attention.dense.qzeros": "model-00001-of-00002.safetensors",
350
+ "gpt_neox.layers.21.attention.dense.scales": "model-00001-of-00002.safetensors",
351
+ "gpt_neox.layers.21.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
352
+ "gpt_neox.layers.21.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
353
+ "gpt_neox.layers.21.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
354
+ "gpt_neox.layers.21.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
355
+ "gpt_neox.layers.21.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
356
+ "gpt_neox.layers.21.input_layernorm.bias": "model-00001-of-00002.safetensors",
357
+ "gpt_neox.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
358
+ "gpt_neox.layers.21.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
359
+ "gpt_neox.layers.21.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
360
+ "gpt_neox.layers.21.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
361
+ "gpt_neox.layers.21.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
362
+ "gpt_neox.layers.21.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
363
+ "gpt_neox.layers.21.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
364
+ "gpt_neox.layers.21.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
365
+ "gpt_neox.layers.21.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
366
+ "gpt_neox.layers.21.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
367
+ "gpt_neox.layers.21.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
368
+ "gpt_neox.layers.21.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
369
+ "gpt_neox.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
370
+ "gpt_neox.layers.22.attention.dense.bias": "model-00001-of-00002.safetensors",
371
+ "gpt_neox.layers.22.attention.dense.g_idx": "model-00001-of-00002.safetensors",
372
+ "gpt_neox.layers.22.attention.dense.qweight": "model-00001-of-00002.safetensors",
373
+ "gpt_neox.layers.22.attention.dense.qzeros": "model-00001-of-00002.safetensors",
374
+ "gpt_neox.layers.22.attention.dense.scales": "model-00001-of-00002.safetensors",
375
+ "gpt_neox.layers.22.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
376
+ "gpt_neox.layers.22.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
377
+ "gpt_neox.layers.22.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
378
+ "gpt_neox.layers.22.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
379
+ "gpt_neox.layers.22.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
380
+ "gpt_neox.layers.22.input_layernorm.bias": "model-00001-of-00002.safetensors",
381
+ "gpt_neox.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
382
+ "gpt_neox.layers.22.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
383
+ "gpt_neox.layers.22.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
384
+ "gpt_neox.layers.22.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
385
+ "gpt_neox.layers.22.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
386
+ "gpt_neox.layers.22.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
387
+ "gpt_neox.layers.22.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
388
+ "gpt_neox.layers.22.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
389
+ "gpt_neox.layers.22.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
390
+ "gpt_neox.layers.22.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
391
+ "gpt_neox.layers.22.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
392
+ "gpt_neox.layers.22.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
393
+ "gpt_neox.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
394
+ "gpt_neox.layers.23.attention.dense.bias": "model-00001-of-00002.safetensors",
395
+ "gpt_neox.layers.23.attention.dense.g_idx": "model-00001-of-00002.safetensors",
396
+ "gpt_neox.layers.23.attention.dense.qweight": "model-00001-of-00002.safetensors",
397
+ "gpt_neox.layers.23.attention.dense.qzeros": "model-00001-of-00002.safetensors",
398
+ "gpt_neox.layers.23.attention.dense.scales": "model-00001-of-00002.safetensors",
399
+ "gpt_neox.layers.23.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
400
+ "gpt_neox.layers.23.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
401
+ "gpt_neox.layers.23.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
402
+ "gpt_neox.layers.23.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
403
+ "gpt_neox.layers.23.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
404
+ "gpt_neox.layers.23.input_layernorm.bias": "model-00001-of-00002.safetensors",
405
+ "gpt_neox.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
406
+ "gpt_neox.layers.23.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
407
+ "gpt_neox.layers.23.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
408
+ "gpt_neox.layers.23.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
409
+ "gpt_neox.layers.23.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
410
+ "gpt_neox.layers.23.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
411
+ "gpt_neox.layers.23.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
412
+ "gpt_neox.layers.23.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
413
+ "gpt_neox.layers.23.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
414
+ "gpt_neox.layers.23.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
415
+ "gpt_neox.layers.23.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
416
+ "gpt_neox.layers.23.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
417
+ "gpt_neox.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
418
+ "gpt_neox.layers.24.attention.dense.bias": "model-00002-of-00002.safetensors",
419
+ "gpt_neox.layers.24.attention.dense.g_idx": "model-00002-of-00002.safetensors",
420
+ "gpt_neox.layers.24.attention.dense.qweight": "model-00002-of-00002.safetensors",
421
+ "gpt_neox.layers.24.attention.dense.qzeros": "model-00002-of-00002.safetensors",
422
+ "gpt_neox.layers.24.attention.dense.scales": "model-00002-of-00002.safetensors",
423
+ "gpt_neox.layers.24.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
424
+ "gpt_neox.layers.24.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
425
+ "gpt_neox.layers.24.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
426
+ "gpt_neox.layers.24.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
427
+ "gpt_neox.layers.24.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
428
+ "gpt_neox.layers.24.input_layernorm.bias": "model-00001-of-00002.safetensors",
429
+ "gpt_neox.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
430
+ "gpt_neox.layers.24.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
431
+ "gpt_neox.layers.24.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
432
+ "gpt_neox.layers.24.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
433
+ "gpt_neox.layers.24.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
434
+ "gpt_neox.layers.24.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
435
+ "gpt_neox.layers.24.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
436
+ "gpt_neox.layers.24.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
437
+ "gpt_neox.layers.24.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
438
+ "gpt_neox.layers.24.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
439
+ "gpt_neox.layers.24.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
440
+ "gpt_neox.layers.24.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
441
+ "gpt_neox.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
442
+ "gpt_neox.layers.25.attention.dense.bias": "model-00002-of-00002.safetensors",
443
+ "gpt_neox.layers.25.attention.dense.g_idx": "model-00002-of-00002.safetensors",
444
+ "gpt_neox.layers.25.attention.dense.qweight": "model-00002-of-00002.safetensors",
445
+ "gpt_neox.layers.25.attention.dense.qzeros": "model-00002-of-00002.safetensors",
446
+ "gpt_neox.layers.25.attention.dense.scales": "model-00002-of-00002.safetensors",
447
+ "gpt_neox.layers.25.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
448
+ "gpt_neox.layers.25.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
449
+ "gpt_neox.layers.25.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
450
+ "gpt_neox.layers.25.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
451
+ "gpt_neox.layers.25.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
452
+ "gpt_neox.layers.25.input_layernorm.bias": "model-00002-of-00002.safetensors",
453
+ "gpt_neox.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors",
454
+ "gpt_neox.layers.25.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
455
+ "gpt_neox.layers.25.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
456
+ "gpt_neox.layers.25.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
457
+ "gpt_neox.layers.25.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
458
+ "gpt_neox.layers.25.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
459
+ "gpt_neox.layers.25.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
460
+ "gpt_neox.layers.25.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
461
+ "gpt_neox.layers.25.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
462
+ "gpt_neox.layers.25.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
463
+ "gpt_neox.layers.25.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
464
+ "gpt_neox.layers.25.post_attention_layernorm.bias": "model-00002-of-00002.safetensors",
465
+ "gpt_neox.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
466
+ "gpt_neox.layers.26.attention.dense.bias": "model-00002-of-00002.safetensors",
467
+ "gpt_neox.layers.26.attention.dense.g_idx": "model-00002-of-00002.safetensors",
468
+ "gpt_neox.layers.26.attention.dense.qweight": "model-00002-of-00002.safetensors",
469
+ "gpt_neox.layers.26.attention.dense.qzeros": "model-00002-of-00002.safetensors",
470
+ "gpt_neox.layers.26.attention.dense.scales": "model-00002-of-00002.safetensors",
471
+ "gpt_neox.layers.26.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
472
+ "gpt_neox.layers.26.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
473
+ "gpt_neox.layers.26.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
474
+ "gpt_neox.layers.26.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
475
+ "gpt_neox.layers.26.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
476
+ "gpt_neox.layers.26.input_layernorm.bias": "model-00002-of-00002.safetensors",
477
+ "gpt_neox.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors",
478
+ "gpt_neox.layers.26.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
479
+ "gpt_neox.layers.26.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
480
+ "gpt_neox.layers.26.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
481
+ "gpt_neox.layers.26.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
482
+ "gpt_neox.layers.26.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
483
+ "gpt_neox.layers.26.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
484
+ "gpt_neox.layers.26.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
485
+ "gpt_neox.layers.26.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
486
+ "gpt_neox.layers.26.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
487
+ "gpt_neox.layers.26.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
488
+ "gpt_neox.layers.26.post_attention_layernorm.bias": "model-00002-of-00002.safetensors",
489
+ "gpt_neox.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
490
+ "gpt_neox.layers.27.attention.dense.bias": "model-00002-of-00002.safetensors",
491
+ "gpt_neox.layers.27.attention.dense.g_idx": "model-00002-of-00002.safetensors",
492
+ "gpt_neox.layers.27.attention.dense.qweight": "model-00002-of-00002.safetensors",
493
+ "gpt_neox.layers.27.attention.dense.qzeros": "model-00002-of-00002.safetensors",
494
+ "gpt_neox.layers.27.attention.dense.scales": "model-00002-of-00002.safetensors",
495
+ "gpt_neox.layers.27.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
496
+ "gpt_neox.layers.27.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
497
+ "gpt_neox.layers.27.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
498
+ "gpt_neox.layers.27.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
499
+ "gpt_neox.layers.27.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
500
+ "gpt_neox.layers.27.input_layernorm.bias": "model-00002-of-00002.safetensors",
501
+ "gpt_neox.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors",
502
+ "gpt_neox.layers.27.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
503
+ "gpt_neox.layers.27.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
504
+ "gpt_neox.layers.27.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
505
+ "gpt_neox.layers.27.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
506
+ "gpt_neox.layers.27.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
507
+ "gpt_neox.layers.27.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
508
+ "gpt_neox.layers.27.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
509
+ "gpt_neox.layers.27.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
510
+ "gpt_neox.layers.27.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
511
+ "gpt_neox.layers.27.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
512
+ "gpt_neox.layers.27.post_attention_layernorm.bias": "model-00002-of-00002.safetensors",
513
+ "gpt_neox.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
514
+ "gpt_neox.layers.28.attention.dense.bias": "model-00002-of-00002.safetensors",
515
+ "gpt_neox.layers.28.attention.dense.g_idx": "model-00002-of-00002.safetensors",
516
+ "gpt_neox.layers.28.attention.dense.qweight": "model-00002-of-00002.safetensors",
517
+ "gpt_neox.layers.28.attention.dense.qzeros": "model-00002-of-00002.safetensors",
518
+ "gpt_neox.layers.28.attention.dense.scales": "model-00002-of-00002.safetensors",
519
+ "gpt_neox.layers.28.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
520
+ "gpt_neox.layers.28.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
521
+ "gpt_neox.layers.28.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
522
+ "gpt_neox.layers.28.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
523
+ "gpt_neox.layers.28.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
524
+ "gpt_neox.layers.28.input_layernorm.bias": "model-00002-of-00002.safetensors",
525
+ "gpt_neox.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors",
526
+ "gpt_neox.layers.28.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
527
+ "gpt_neox.layers.28.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
528
+ "gpt_neox.layers.28.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
529
+ "gpt_neox.layers.28.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
530
+ "gpt_neox.layers.28.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
531
+ "gpt_neox.layers.28.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
532
+ "gpt_neox.layers.28.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
533
+ "gpt_neox.layers.28.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
534
+ "gpt_neox.layers.28.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
535
+ "gpt_neox.layers.28.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
536
+ "gpt_neox.layers.28.post_attention_layernorm.bias": "model-00002-of-00002.safetensors",
537
+ "gpt_neox.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
538
+ "gpt_neox.layers.29.attention.dense.bias": "model-00002-of-00002.safetensors",
539
+ "gpt_neox.layers.29.attention.dense.g_idx": "model-00002-of-00002.safetensors",
540
+ "gpt_neox.layers.29.attention.dense.qweight": "model-00002-of-00002.safetensors",
541
+ "gpt_neox.layers.29.attention.dense.qzeros": "model-00002-of-00002.safetensors",
542
+ "gpt_neox.layers.29.attention.dense.scales": "model-00002-of-00002.safetensors",
543
+ "gpt_neox.layers.29.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
544
+ "gpt_neox.layers.29.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
545
+ "gpt_neox.layers.29.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
546
+ "gpt_neox.layers.29.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
547
+ "gpt_neox.layers.29.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
548
+ "gpt_neox.layers.29.input_layernorm.bias": "model-00002-of-00002.safetensors",
549
+ "gpt_neox.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors",
550
+ "gpt_neox.layers.29.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
551
+ "gpt_neox.layers.29.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
552
+ "gpt_neox.layers.29.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
553
+ "gpt_neox.layers.29.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
554
+ "gpt_neox.layers.29.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
555
+ "gpt_neox.layers.29.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
556
+ "gpt_neox.layers.29.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
557
+ "gpt_neox.layers.29.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
558
+ "gpt_neox.layers.29.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
559
+ "gpt_neox.layers.29.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
560
+ "gpt_neox.layers.29.post_attention_layernorm.bias": "model-00002-of-00002.safetensors",
561
+ "gpt_neox.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
562
+ "gpt_neox.layers.3.attention.dense.bias": "model-00001-of-00002.safetensors",
563
+ "gpt_neox.layers.3.attention.dense.g_idx": "model-00001-of-00002.safetensors",
564
+ "gpt_neox.layers.3.attention.dense.qweight": "model-00001-of-00002.safetensors",
565
+ "gpt_neox.layers.3.attention.dense.qzeros": "model-00001-of-00002.safetensors",
566
+ "gpt_neox.layers.3.attention.dense.scales": "model-00001-of-00002.safetensors",
567
+ "gpt_neox.layers.3.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
568
+ "gpt_neox.layers.3.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
569
+ "gpt_neox.layers.3.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
570
+ "gpt_neox.layers.3.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
571
+ "gpt_neox.layers.3.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
572
+ "gpt_neox.layers.3.input_layernorm.bias": "model-00001-of-00002.safetensors",
573
+ "gpt_neox.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
574
+ "gpt_neox.layers.3.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
575
+ "gpt_neox.layers.3.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
576
+ "gpt_neox.layers.3.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
577
+ "gpt_neox.layers.3.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
578
+ "gpt_neox.layers.3.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
579
+ "gpt_neox.layers.3.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
580
+ "gpt_neox.layers.3.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
581
+ "gpt_neox.layers.3.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
582
+ "gpt_neox.layers.3.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
583
+ "gpt_neox.layers.3.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
584
+ "gpt_neox.layers.3.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
585
+ "gpt_neox.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
586
+ "gpt_neox.layers.30.attention.dense.bias": "model-00002-of-00002.safetensors",
587
+ "gpt_neox.layers.30.attention.dense.g_idx": "model-00002-of-00002.safetensors",
588
+ "gpt_neox.layers.30.attention.dense.qweight": "model-00002-of-00002.safetensors",
589
+ "gpt_neox.layers.30.attention.dense.qzeros": "model-00002-of-00002.safetensors",
590
+ "gpt_neox.layers.30.attention.dense.scales": "model-00002-of-00002.safetensors",
591
+ "gpt_neox.layers.30.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
592
+ "gpt_neox.layers.30.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
593
+ "gpt_neox.layers.30.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
594
+ "gpt_neox.layers.30.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
595
+ "gpt_neox.layers.30.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
596
+ "gpt_neox.layers.30.input_layernorm.bias": "model-00002-of-00002.safetensors",
597
+ "gpt_neox.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors",
598
+ "gpt_neox.layers.30.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
599
+ "gpt_neox.layers.30.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
600
+ "gpt_neox.layers.30.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
601
+ "gpt_neox.layers.30.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
602
+ "gpt_neox.layers.30.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
603
+ "gpt_neox.layers.30.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
604
+ "gpt_neox.layers.30.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
605
+ "gpt_neox.layers.30.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
606
+ "gpt_neox.layers.30.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
607
+ "gpt_neox.layers.30.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
608
+ "gpt_neox.layers.30.post_attention_layernorm.bias": "model-00002-of-00002.safetensors",
609
+ "gpt_neox.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
610
+ "gpt_neox.layers.31.attention.dense.bias": "model-00002-of-00002.safetensors",
611
+ "gpt_neox.layers.31.attention.dense.g_idx": "model-00002-of-00002.safetensors",
612
+ "gpt_neox.layers.31.attention.dense.qweight": "model-00002-of-00002.safetensors",
613
+ "gpt_neox.layers.31.attention.dense.qzeros": "model-00002-of-00002.safetensors",
614
+ "gpt_neox.layers.31.attention.dense.scales": "model-00002-of-00002.safetensors",
615
+ "gpt_neox.layers.31.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
616
+ "gpt_neox.layers.31.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
617
+ "gpt_neox.layers.31.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
618
+ "gpt_neox.layers.31.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
619
+ "gpt_neox.layers.31.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
620
+ "gpt_neox.layers.31.input_layernorm.bias": "model-00002-of-00002.safetensors",
621
+ "gpt_neox.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors",
622
+ "gpt_neox.layers.31.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
623
+ "gpt_neox.layers.31.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
624
+ "gpt_neox.layers.31.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
625
+ "gpt_neox.layers.31.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
626
+ "gpt_neox.layers.31.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
627
+ "gpt_neox.layers.31.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
628
+ "gpt_neox.layers.31.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
629
+ "gpt_neox.layers.31.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
630
+ "gpt_neox.layers.31.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
631
+ "gpt_neox.layers.31.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
632
+ "gpt_neox.layers.31.post_attention_layernorm.bias": "model-00002-of-00002.safetensors",
633
+ "gpt_neox.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
634
+ "gpt_neox.layers.32.attention.dense.bias": "model-00002-of-00002.safetensors",
635
+ "gpt_neox.layers.32.attention.dense.g_idx": "model-00002-of-00002.safetensors",
636
+ "gpt_neox.layers.32.attention.dense.qweight": "model-00002-of-00002.safetensors",
637
+ "gpt_neox.layers.32.attention.dense.qzeros": "model-00002-of-00002.safetensors",
638
+ "gpt_neox.layers.32.attention.dense.scales": "model-00002-of-00002.safetensors",
639
+ "gpt_neox.layers.32.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
640
+ "gpt_neox.layers.32.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
641
+ "gpt_neox.layers.32.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
642
+ "gpt_neox.layers.32.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
643
+ "gpt_neox.layers.32.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
644
+ "gpt_neox.layers.32.input_layernorm.bias": "model-00002-of-00002.safetensors",
645
+ "gpt_neox.layers.32.input_layernorm.weight": "model-00002-of-00002.safetensors",
646
+ "gpt_neox.layers.32.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
647
+ "gpt_neox.layers.32.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
648
+ "gpt_neox.layers.32.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
649
+ "gpt_neox.layers.32.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
650
+ "gpt_neox.layers.32.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
651
+ "gpt_neox.layers.32.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
652
+ "gpt_neox.layers.32.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
653
+ "gpt_neox.layers.32.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
654
+ "gpt_neox.layers.32.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
655
+ "gpt_neox.layers.32.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
656
+ "gpt_neox.layers.32.post_attention_layernorm.bias": "model-00002-of-00002.safetensors",
657
+ "gpt_neox.layers.32.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
658
+ "gpt_neox.layers.33.attention.dense.bias": "model-00002-of-00002.safetensors",
659
+ "gpt_neox.layers.33.attention.dense.g_idx": "model-00002-of-00002.safetensors",
660
+ "gpt_neox.layers.33.attention.dense.qweight": "model-00002-of-00002.safetensors",
661
+ "gpt_neox.layers.33.attention.dense.qzeros": "model-00002-of-00002.safetensors",
662
+ "gpt_neox.layers.33.attention.dense.scales": "model-00002-of-00002.safetensors",
663
+ "gpt_neox.layers.33.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
664
+ "gpt_neox.layers.33.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
665
+ "gpt_neox.layers.33.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
666
+ "gpt_neox.layers.33.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
667
+ "gpt_neox.layers.33.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
668
+ "gpt_neox.layers.33.input_layernorm.bias": "model-00002-of-00002.safetensors",
669
+ "gpt_neox.layers.33.input_layernorm.weight": "model-00002-of-00002.safetensors",
670
+ "gpt_neox.layers.33.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
671
+ "gpt_neox.layers.33.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
672
+ "gpt_neox.layers.33.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
673
+ "gpt_neox.layers.33.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
674
+ "gpt_neox.layers.33.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
675
+ "gpt_neox.layers.33.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
676
+ "gpt_neox.layers.33.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
677
+ "gpt_neox.layers.33.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
678
+ "gpt_neox.layers.33.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
679
+ "gpt_neox.layers.33.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
680
+ "gpt_neox.layers.33.post_attention_layernorm.bias": "model-00002-of-00002.safetensors",
681
+ "gpt_neox.layers.33.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
682
+ "gpt_neox.layers.34.attention.dense.bias": "model-00002-of-00002.safetensors",
683
+ "gpt_neox.layers.34.attention.dense.g_idx": "model-00002-of-00002.safetensors",
684
+ "gpt_neox.layers.34.attention.dense.qweight": "model-00002-of-00002.safetensors",
685
+ "gpt_neox.layers.34.attention.dense.qzeros": "model-00002-of-00002.safetensors",
686
+ "gpt_neox.layers.34.attention.dense.scales": "model-00002-of-00002.safetensors",
687
+ "gpt_neox.layers.34.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
688
+ "gpt_neox.layers.34.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
689
+ "gpt_neox.layers.34.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
690
+ "gpt_neox.layers.34.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
691
+ "gpt_neox.layers.34.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
692
+ "gpt_neox.layers.34.input_layernorm.bias": "model-00002-of-00002.safetensors",
693
+ "gpt_neox.layers.34.input_layernorm.weight": "model-00002-of-00002.safetensors",
694
+ "gpt_neox.layers.34.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
695
+ "gpt_neox.layers.34.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
696
+ "gpt_neox.layers.34.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
697
+ "gpt_neox.layers.34.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
698
+ "gpt_neox.layers.34.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
699
+ "gpt_neox.layers.34.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
700
+ "gpt_neox.layers.34.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
701
+ "gpt_neox.layers.34.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
702
+ "gpt_neox.layers.34.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
703
+ "gpt_neox.layers.34.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
704
+ "gpt_neox.layers.34.post_attention_layernorm.bias": "model-00002-of-00002.safetensors",
705
+ "gpt_neox.layers.34.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
706
+ "gpt_neox.layers.35.attention.dense.bias": "model-00002-of-00002.safetensors",
707
+ "gpt_neox.layers.35.attention.dense.g_idx": "model-00002-of-00002.safetensors",
708
+ "gpt_neox.layers.35.attention.dense.qweight": "model-00002-of-00002.safetensors",
709
+ "gpt_neox.layers.35.attention.dense.qzeros": "model-00002-of-00002.safetensors",
710
+ "gpt_neox.layers.35.attention.dense.scales": "model-00002-of-00002.safetensors",
711
+ "gpt_neox.layers.35.attention.query_key_value.bias": "model-00002-of-00002.safetensors",
712
+ "gpt_neox.layers.35.attention.query_key_value.g_idx": "model-00002-of-00002.safetensors",
713
+ "gpt_neox.layers.35.attention.query_key_value.qweight": "model-00002-of-00002.safetensors",
714
+ "gpt_neox.layers.35.attention.query_key_value.qzeros": "model-00002-of-00002.safetensors",
715
+ "gpt_neox.layers.35.attention.query_key_value.scales": "model-00002-of-00002.safetensors",
716
+ "gpt_neox.layers.35.input_layernorm.bias": "model-00002-of-00002.safetensors",
717
+ "gpt_neox.layers.35.input_layernorm.weight": "model-00002-of-00002.safetensors",
718
+ "gpt_neox.layers.35.mlp.dense_4h_to_h.bias": "model-00002-of-00002.safetensors",
719
+ "gpt_neox.layers.35.mlp.dense_4h_to_h.g_idx": "model-00002-of-00002.safetensors",
720
+ "gpt_neox.layers.35.mlp.dense_4h_to_h.qweight": "model-00002-of-00002.safetensors",
721
+ "gpt_neox.layers.35.mlp.dense_4h_to_h.qzeros": "model-00002-of-00002.safetensors",
722
+ "gpt_neox.layers.35.mlp.dense_4h_to_h.scales": "model-00002-of-00002.safetensors",
723
+ "gpt_neox.layers.35.mlp.dense_h_to_4h.bias": "model-00002-of-00002.safetensors",
724
+ "gpt_neox.layers.35.mlp.dense_h_to_4h.g_idx": "model-00002-of-00002.safetensors",
725
+ "gpt_neox.layers.35.mlp.dense_h_to_4h.qweight": "model-00002-of-00002.safetensors",
726
+ "gpt_neox.layers.35.mlp.dense_h_to_4h.qzeros": "model-00002-of-00002.safetensors",
727
+ "gpt_neox.layers.35.mlp.dense_h_to_4h.scales": "model-00002-of-00002.safetensors",
728
+ "gpt_neox.layers.35.post_attention_layernorm.bias": "model-00002-of-00002.safetensors",
729
+ "gpt_neox.layers.35.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
730
+ "gpt_neox.layers.4.attention.dense.bias": "model-00001-of-00002.safetensors",
731
+ "gpt_neox.layers.4.attention.dense.g_idx": "model-00001-of-00002.safetensors",
732
+ "gpt_neox.layers.4.attention.dense.qweight": "model-00001-of-00002.safetensors",
733
+ "gpt_neox.layers.4.attention.dense.qzeros": "model-00001-of-00002.safetensors",
734
+ "gpt_neox.layers.4.attention.dense.scales": "model-00001-of-00002.safetensors",
735
+ "gpt_neox.layers.4.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
736
+ "gpt_neox.layers.4.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
737
+ "gpt_neox.layers.4.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
738
+ "gpt_neox.layers.4.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
739
+ "gpt_neox.layers.4.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
740
+ "gpt_neox.layers.4.input_layernorm.bias": "model-00001-of-00002.safetensors",
741
+ "gpt_neox.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
742
+ "gpt_neox.layers.4.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
743
+ "gpt_neox.layers.4.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
744
+ "gpt_neox.layers.4.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
745
+ "gpt_neox.layers.4.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
746
+ "gpt_neox.layers.4.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
747
+ "gpt_neox.layers.4.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
748
+ "gpt_neox.layers.4.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
749
+ "gpt_neox.layers.4.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
750
+ "gpt_neox.layers.4.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
751
+ "gpt_neox.layers.4.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
752
+ "gpt_neox.layers.4.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
753
+ "gpt_neox.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
754
+ "gpt_neox.layers.5.attention.dense.bias": "model-00001-of-00002.safetensors",
755
+ "gpt_neox.layers.5.attention.dense.g_idx": "model-00001-of-00002.safetensors",
756
+ "gpt_neox.layers.5.attention.dense.qweight": "model-00001-of-00002.safetensors",
757
+ "gpt_neox.layers.5.attention.dense.qzeros": "model-00001-of-00002.safetensors",
758
+ "gpt_neox.layers.5.attention.dense.scales": "model-00001-of-00002.safetensors",
759
+ "gpt_neox.layers.5.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
760
+ "gpt_neox.layers.5.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
761
+ "gpt_neox.layers.5.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
762
+ "gpt_neox.layers.5.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
763
+ "gpt_neox.layers.5.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
764
+ "gpt_neox.layers.5.input_layernorm.bias": "model-00001-of-00002.safetensors",
765
+ "gpt_neox.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
766
+ "gpt_neox.layers.5.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
767
+ "gpt_neox.layers.5.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
768
+ "gpt_neox.layers.5.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
769
+ "gpt_neox.layers.5.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
770
+ "gpt_neox.layers.5.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
771
+ "gpt_neox.layers.5.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
772
+ "gpt_neox.layers.5.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
773
+ "gpt_neox.layers.5.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
774
+ "gpt_neox.layers.5.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
775
+ "gpt_neox.layers.5.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
776
+ "gpt_neox.layers.5.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
777
+ "gpt_neox.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
778
+ "gpt_neox.layers.6.attention.dense.bias": "model-00001-of-00002.safetensors",
779
+ "gpt_neox.layers.6.attention.dense.g_idx": "model-00001-of-00002.safetensors",
780
+ "gpt_neox.layers.6.attention.dense.qweight": "model-00001-of-00002.safetensors",
781
+ "gpt_neox.layers.6.attention.dense.qzeros": "model-00001-of-00002.safetensors",
782
+ "gpt_neox.layers.6.attention.dense.scales": "model-00001-of-00002.safetensors",
783
+ "gpt_neox.layers.6.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
784
+ "gpt_neox.layers.6.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
785
+ "gpt_neox.layers.6.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
786
+ "gpt_neox.layers.6.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
787
+ "gpt_neox.layers.6.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
788
+ "gpt_neox.layers.6.input_layernorm.bias": "model-00001-of-00002.safetensors",
789
+ "gpt_neox.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
790
+ "gpt_neox.layers.6.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
791
+ "gpt_neox.layers.6.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
792
+ "gpt_neox.layers.6.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
793
+ "gpt_neox.layers.6.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
794
+ "gpt_neox.layers.6.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
795
+ "gpt_neox.layers.6.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
796
+ "gpt_neox.layers.6.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
797
+ "gpt_neox.layers.6.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
798
+ "gpt_neox.layers.6.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
799
+ "gpt_neox.layers.6.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
800
+ "gpt_neox.layers.6.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
801
+ "gpt_neox.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
802
+ "gpt_neox.layers.7.attention.dense.bias": "model-00001-of-00002.safetensors",
803
+ "gpt_neox.layers.7.attention.dense.g_idx": "model-00001-of-00002.safetensors",
804
+ "gpt_neox.layers.7.attention.dense.qweight": "model-00001-of-00002.safetensors",
805
+ "gpt_neox.layers.7.attention.dense.qzeros": "model-00001-of-00002.safetensors",
806
+ "gpt_neox.layers.7.attention.dense.scales": "model-00001-of-00002.safetensors",
807
+ "gpt_neox.layers.7.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
808
+ "gpt_neox.layers.7.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
809
+ "gpt_neox.layers.7.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
810
+ "gpt_neox.layers.7.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
811
+ "gpt_neox.layers.7.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
812
+ "gpt_neox.layers.7.input_layernorm.bias": "model-00001-of-00002.safetensors",
813
+ "gpt_neox.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
814
+ "gpt_neox.layers.7.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
815
+ "gpt_neox.layers.7.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
816
+ "gpt_neox.layers.7.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
817
+ "gpt_neox.layers.7.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
818
+ "gpt_neox.layers.7.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
819
+ "gpt_neox.layers.7.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
820
+ "gpt_neox.layers.7.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
821
+ "gpt_neox.layers.7.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
822
+ "gpt_neox.layers.7.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
823
+ "gpt_neox.layers.7.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
824
+ "gpt_neox.layers.7.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
825
+ "gpt_neox.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
826
+ "gpt_neox.layers.8.attention.dense.bias": "model-00001-of-00002.safetensors",
827
+ "gpt_neox.layers.8.attention.dense.g_idx": "model-00001-of-00002.safetensors",
828
+ "gpt_neox.layers.8.attention.dense.qweight": "model-00001-of-00002.safetensors",
829
+ "gpt_neox.layers.8.attention.dense.qzeros": "model-00001-of-00002.safetensors",
830
+ "gpt_neox.layers.8.attention.dense.scales": "model-00001-of-00002.safetensors",
831
+ "gpt_neox.layers.8.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
832
+ "gpt_neox.layers.8.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
833
+ "gpt_neox.layers.8.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
834
+ "gpt_neox.layers.8.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
835
+ "gpt_neox.layers.8.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
836
+ "gpt_neox.layers.8.input_layernorm.bias": "model-00001-of-00002.safetensors",
837
+ "gpt_neox.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
838
+ "gpt_neox.layers.8.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
839
+ "gpt_neox.layers.8.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
840
+ "gpt_neox.layers.8.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
841
+ "gpt_neox.layers.8.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
842
+ "gpt_neox.layers.8.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
843
+ "gpt_neox.layers.8.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
844
+ "gpt_neox.layers.8.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
845
+ "gpt_neox.layers.8.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
846
+ "gpt_neox.layers.8.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
847
+ "gpt_neox.layers.8.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
848
+ "gpt_neox.layers.8.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
849
+ "gpt_neox.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
850
+ "gpt_neox.layers.9.attention.dense.bias": "model-00001-of-00002.safetensors",
851
+ "gpt_neox.layers.9.attention.dense.g_idx": "model-00001-of-00002.safetensors",
852
+ "gpt_neox.layers.9.attention.dense.qweight": "model-00001-of-00002.safetensors",
853
+ "gpt_neox.layers.9.attention.dense.qzeros": "model-00001-of-00002.safetensors",
854
+ "gpt_neox.layers.9.attention.dense.scales": "model-00001-of-00002.safetensors",
855
+ "gpt_neox.layers.9.attention.query_key_value.bias": "model-00001-of-00002.safetensors",
856
+ "gpt_neox.layers.9.attention.query_key_value.g_idx": "model-00001-of-00002.safetensors",
857
+ "gpt_neox.layers.9.attention.query_key_value.qweight": "model-00001-of-00002.safetensors",
858
+ "gpt_neox.layers.9.attention.query_key_value.qzeros": "model-00001-of-00002.safetensors",
859
+ "gpt_neox.layers.9.attention.query_key_value.scales": "model-00001-of-00002.safetensors",
860
+ "gpt_neox.layers.9.input_layernorm.bias": "model-00001-of-00002.safetensors",
861
+ "gpt_neox.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
862
+ "gpt_neox.layers.9.mlp.dense_4h_to_h.bias": "model-00001-of-00002.safetensors",
863
+ "gpt_neox.layers.9.mlp.dense_4h_to_h.g_idx": "model-00001-of-00002.safetensors",
864
+ "gpt_neox.layers.9.mlp.dense_4h_to_h.qweight": "model-00001-of-00002.safetensors",
865
+ "gpt_neox.layers.9.mlp.dense_4h_to_h.qzeros": "model-00001-of-00002.safetensors",
866
+ "gpt_neox.layers.9.mlp.dense_4h_to_h.scales": "model-00001-of-00002.safetensors",
867
+ "gpt_neox.layers.9.mlp.dense_h_to_4h.bias": "model-00001-of-00002.safetensors",
868
+ "gpt_neox.layers.9.mlp.dense_h_to_4h.g_idx": "model-00001-of-00002.safetensors",
869
+ "gpt_neox.layers.9.mlp.dense_h_to_4h.qweight": "model-00001-of-00002.safetensors",
870
+ "gpt_neox.layers.9.mlp.dense_h_to_4h.qzeros": "model-00001-of-00002.safetensors",
871
+ "gpt_neox.layers.9.mlp.dense_h_to_4h.scales": "model-00001-of-00002.safetensors",
872
+ "gpt_neox.layers.9.post_attention_layernorm.bias": "model-00001-of-00002.safetensors",
873
+ "gpt_neox.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors"
874
+ }
875
+ }
quantization_config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bits": 4,
3
+ "group_size": 128,
4
+ "sym": false,
5
+ "data_type": "int",
6
+ "enable_quanted_input": true,
7
+ "enable_minmax_tuning": true,
8
+ "seqlen": 512,
9
+ "train_bs": 4,
10
+ "scale_dtype": "torch.float16",
11
+ "lr": 0.005,
12
+ "minmax_lr": 0.005,
13
+ "gradient_accumulate_steps": 1,
14
+ "iters": 200,
15
+ "amp": false,
16
+ "nsamples": 128,
17
+ "low_gpu_mem_usage": false,
18
+ "quant_block_list": null,
19
+ "enable_norm_bias_tuning": false,
20
+ "dataset": "NeelNanda/pile-10k",
21
+ "autoround_version": "0.4.0.dev",
22
+ "quant_method": "intel/auto-round",
23
+ "backend": "auto_round:exllamav2"
24
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|endoftext|>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<|endoftext|>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "unk_token": {
17
+ "content": "<|endoftext|>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ }
23
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,214 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_eos_token": false,
4
+ "add_prefix_space": false,
5
+ "added_tokens_decoder": {
6
+ "0": {
7
+ "content": "<|endoftext|>",
8
+ "lstrip": false,
9
+ "normalized": false,
10
+ "rstrip": false,
11
+ "single_word": false,
12
+ "special": true
13
+ },
14
+ "1": {
15
+ "content": "<|padding|>",
16
+ "lstrip": false,
17
+ "normalized": false,
18
+ "rstrip": false,
19
+ "single_word": false,
20
+ "special": true
21
+ },
22
+ "50254": {
23
+ "content": " ",
24
+ "lstrip": false,
25
+ "normalized": true,
26
+ "rstrip": false,
27
+ "single_word": false,
28
+ "special": false
29
+ },
30
+ "50255": {
31
+ "content": " ",
32
+ "lstrip": false,
33
+ "normalized": true,
34
+ "rstrip": false,
35
+ "single_word": false,
36
+ "special": false
37
+ },
38
+ "50256": {
39
+ "content": " ",
40
+ "lstrip": false,
41
+ "normalized": true,
42
+ "rstrip": false,
43
+ "single_word": false,
44
+ "special": false
45
+ },
46
+ "50257": {
47
+ "content": " ",
48
+ "lstrip": false,
49
+ "normalized": true,
50
+ "rstrip": false,
51
+ "single_word": false,
52
+ "special": false
53
+ },
54
+ "50258": {
55
+ "content": " ",
56
+ "lstrip": false,
57
+ "normalized": true,
58
+ "rstrip": false,
59
+ "single_word": false,
60
+ "special": false
61
+ },
62
+ "50259": {
63
+ "content": " ",
64
+ "lstrip": false,
65
+ "normalized": true,
66
+ "rstrip": false,
67
+ "single_word": false,
68
+ "special": false
69
+ },
70
+ "50260": {
71
+ "content": " ",
72
+ "lstrip": false,
73
+ "normalized": true,
74
+ "rstrip": false,
75
+ "single_word": false,
76
+ "special": false
77
+ },
78
+ "50261": {
79
+ "content": " ",
80
+ "lstrip": false,
81
+ "normalized": true,
82
+ "rstrip": false,
83
+ "single_word": false,
84
+ "special": false
85
+ },
86
+ "50262": {
87
+ "content": " ",
88
+ "lstrip": false,
89
+ "normalized": true,
90
+ "rstrip": false,
91
+ "single_word": false,
92
+ "special": false
93
+ },
94
+ "50263": {
95
+ "content": " ",
96
+ "lstrip": false,
97
+ "normalized": true,
98
+ "rstrip": false,
99
+ "single_word": false,
100
+ "special": false
101
+ },
102
+ "50264": {
103
+ "content": " ",
104
+ "lstrip": false,
105
+ "normalized": true,
106
+ "rstrip": false,
107
+ "single_word": false,
108
+ "special": false
109
+ },
110
+ "50265": {
111
+ "content": " ",
112
+ "lstrip": false,
113
+ "normalized": true,
114
+ "rstrip": false,
115
+ "single_word": false,
116
+ "special": false
117
+ },
118
+ "50266": {
119
+ "content": " ",
120
+ "lstrip": false,
121
+ "normalized": true,
122
+ "rstrip": false,
123
+ "single_word": false,
124
+ "special": false
125
+ },
126
+ "50267": {
127
+ "content": " ",
128
+ "lstrip": false,
129
+ "normalized": true,
130
+ "rstrip": false,
131
+ "single_word": false,
132
+ "special": false
133
+ },
134
+ "50268": {
135
+ "content": " ",
136
+ "lstrip": false,
137
+ "normalized": true,
138
+ "rstrip": false,
139
+ "single_word": false,
140
+ "special": false
141
+ },
142
+ "50269": {
143
+ "content": " ",
144
+ "lstrip": false,
145
+ "normalized": true,
146
+ "rstrip": false,
147
+ "single_word": false,
148
+ "special": false
149
+ },
150
+ "50270": {
151
+ "content": " ",
152
+ "lstrip": false,
153
+ "normalized": true,
154
+ "rstrip": false,
155
+ "single_word": false,
156
+ "special": false
157
+ },
158
+ "50271": {
159
+ "content": " ",
160
+ "lstrip": false,
161
+ "normalized": true,
162
+ "rstrip": false,
163
+ "single_word": false,
164
+ "special": false
165
+ },
166
+ "50272": {
167
+ "content": " ",
168
+ "lstrip": false,
169
+ "normalized": true,
170
+ "rstrip": false,
171
+ "single_word": false,
172
+ "special": false
173
+ },
174
+ "50273": {
175
+ "content": " ",
176
+ "lstrip": false,
177
+ "normalized": true,
178
+ "rstrip": false,
179
+ "single_word": false,
180
+ "special": false
181
+ },
182
+ "50274": {
183
+ "content": " ",
184
+ "lstrip": false,
185
+ "normalized": true,
186
+ "rstrip": false,
187
+ "single_word": false,
188
+ "special": false
189
+ },
190
+ "50275": {
191
+ "content": " ",
192
+ "lstrip": false,
193
+ "normalized": true,
194
+ "rstrip": false,
195
+ "single_word": false,
196
+ "special": false
197
+ },
198
+ "50276": {
199
+ "content": " ",
200
+ "lstrip": false,
201
+ "normalized": true,
202
+ "rstrip": false,
203
+ "single_word": false,
204
+ "special": false
205
+ }
206
+ },
207
+ "bos_token": "<|endoftext|>",
208
+ "clean_up_tokenization_spaces": false,
209
+ "eos_token": "<|endoftext|>",
210
+ "model_max_length": 1000000000000000019884624838656,
211
+ "pad_token": null,
212
+ "tokenizer_class": "GPTNeoXTokenizer",
213
+ "unk_token": "<|endoftext|>"
214
+ }