jlli commited on
Commit
327693c
1 Parent(s): 9ebbe9e

96f1038ced7f055c819a932e44187c5d5b06bcaa533a933d2c93cd3667df32b9

Browse files
.gitattributes CHANGED
@@ -32,3 +32,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
35
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/nvmedata/jue/bloom_ock_new",
3
+ "apply_residual_connection_post_layernorm": false,
4
+ "architectures": [
5
+ "BloomForCausalLM"
6
+ ],
7
+ "attention_dropout": 0.0,
8
+ "attention_softmax_in_fp32": true,
9
+ "bos_token_id": 1,
10
+ "eos_token_id": 2,
11
+ "hidden_dropout": 0.0,
12
+ "hidden_size": 14336,
13
+ "initializer_range": 0.02,
14
+ "layer_norm_epsilon": 1e-05,
15
+ "masked_softmax_fusion": true,
16
+ "model_type": "bloom",
17
+ "n_head": 112,
18
+ "n_layer": 70,
19
+ "pad_token_id": 3,
20
+ "pretraining_tp": 4,
21
+ "return_dict": false,
22
+ "slow_but_exact": false,
23
+ "transformers_version": "4.25.0",
24
+ "use_cache": false,
25
+ "vocab_size": 250880
26
+ }
pytorch_model.bin.index.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"metadata": {}, "weight_map": {"transformer.word_embeddings.weight": "pytorch_model-01-of-20.bin", "transformer.word_embeddings_layernorm.weight": "pytorch_model-01-of-20.bin", "transformer.word_embeddings_layernorm.bias": "pytorch_model-01-of-20.bin", "transformer.h.0.input_layernorm.weight": "pytorch_model-01-of-20.bin", "transformer.h.0.input_layernorm.bias": "pytorch_model-01-of-20.bin", "transformer.h.0.self_attention.query_key_value.weight": "pytorch_model-01-of-20.bin", "transformer.h.0.self_attention.query_key_value.bias": "pytorch_model-01-of-20.bin", "transformer.h.0.self_attention.dense.weight": "pytorch_model-01-of-20.bin", "transformer.h.0.self_attention.dense.bias": "pytorch_model-01-of-20.bin", "transformer.h.0.post_attention_layernorm.weight": "pytorch_model-01-of-20.bin", "transformer.h.0.post_attention_layernorm.bias": "pytorch_model-01-of-20.bin", "transformer.h.0.mlp.dense_h_to_4h.weight": "pytorch_model-01-of-20.bin", "transformer.h.0.mlp.dense_h_to_4h.bias": "pytorch_model-01-of-20.bin", "transformer.h.0.mlp.dense_4h_to_h.weight": "pytorch_model-01-of-20.bin", "transformer.h.0.mlp.dense_4h_to_h.bias": "pytorch_model-01-of-20.bin", "transformer.h.1.input_layernorm.weight": "pytorch_model-01-of-20.bin", "transformer.h.1.input_layernorm.bias": "pytorch_model-01-of-20.bin", "transformer.h.1.self_attention.query_key_value.weight": "pytorch_model-01-of-20.bin", "transformer.h.1.self_attention.query_key_value.bias": "pytorch_model-01-of-20.bin", "transformer.h.1.self_attention.dense.weight": "pytorch_model-01-of-20.bin", "transformer.h.1.self_attention.dense.bias": "pytorch_model-01-of-20.bin", "transformer.h.1.post_attention_layernorm.weight": "pytorch_model-01-of-20.bin", "transformer.h.1.post_attention_layernorm.bias": "pytorch_model-01-of-20.bin", "transformer.h.1.mlp.dense_h_to_4h.weight": "pytorch_model-01-of-20.bin", "transformer.h.1.mlp.dense_h_to_4h.bias": "pytorch_model-01-of-20.bin", "transformer.h.1.mlp.dense_4h_to_h.weight": "pytorch_model-01-of-20.bin", "transformer.h.1.mlp.dense_4h_to_h.bias": "pytorch_model-01-of-20.bin", "transformer.h.2.input_layernorm.weight": "pytorch_model-01-of-20.bin", "transformer.h.2.input_layernorm.bias": "pytorch_model-01-of-20.bin", "transformer.h.2.self_attention.query_key_value.weight": "pytorch_model-02-of-20.bin", "transformer.h.2.self_attention.query_key_value.bias": "pytorch_model-02-of-20.bin", "transformer.h.2.self_attention.dense.weight": "pytorch_model-02-of-20.bin", "transformer.h.2.self_attention.dense.bias": "pytorch_model-02-of-20.bin", "transformer.h.2.post_attention_layernorm.weight": "pytorch_model-02-of-20.bin", "transformer.h.2.post_attention_layernorm.bias": "pytorch_model-02-of-20.bin", "transformer.h.2.mlp.dense_h_to_4h.weight": "pytorch_model-02-of-20.bin", "transformer.h.2.mlp.dense_h_to_4h.bias": "pytorch_model-02-of-20.bin", "transformer.h.2.mlp.dense_4h_to_h.weight": "pytorch_model-02-of-20.bin", "transformer.h.2.mlp.dense_4h_to_h.bias": "pytorch_model-02-of-20.bin", "transformer.h.3.input_layernorm.weight": "pytorch_model-02-of-20.bin", "transformer.h.3.input_layernorm.bias": "pytorch_model-02-of-20.bin", "transformer.h.3.self_attention.query_key_value.weight": "pytorch_model-02-of-20.bin", "transformer.h.3.self_attention.query_key_value.bias": "pytorch_model-02-of-20.bin", "transformer.h.3.self_attention.dense.weight": "pytorch_model-02-of-20.bin", "transformer.h.3.self_attention.dense.bias": "pytorch_model-02-of-20.bin", "transformer.h.3.post_attention_layernorm.weight": "pytorch_model-02-of-20.bin", "transformer.h.3.post_attention_layernorm.bias": "pytorch_model-02-of-20.bin", "transformer.h.3.mlp.dense_h_to_4h.weight": "pytorch_model-02-of-20.bin", "transformer.h.3.mlp.dense_h_to_4h.bias": "pytorch_model-02-of-20.bin", "transformer.h.3.mlp.dense_4h_to_h.weight": "pytorch_model-02-of-20.bin", "transformer.h.3.mlp.dense_4h_to_h.bias": "pytorch_model-02-of-20.bin", "transformer.h.4.input_layernorm.weight": "pytorch_model-02-of-20.bin", "transformer.h.4.input_layernorm.bias": "pytorch_model-02-of-20.bin", "transformer.h.4.self_attention.query_key_value.weight": "pytorch_model-02-of-20.bin", "transformer.h.4.self_attention.query_key_value.bias": "pytorch_model-02-of-20.bin", "transformer.h.4.self_attention.dense.weight": "pytorch_model-02-of-20.bin", "transformer.h.4.self_attention.dense.bias": "pytorch_model-02-of-20.bin", "transformer.h.4.post_attention_layernorm.weight": "pytorch_model-02-of-20.bin", "transformer.h.4.post_attention_layernorm.bias": "pytorch_model-02-of-20.bin", "transformer.h.4.mlp.dense_h_to_4h.weight": "pytorch_model-02-of-20.bin", "transformer.h.4.mlp.dense_h_to_4h.bias": "pytorch_model-02-of-20.bin", "transformer.h.4.mlp.dense_4h_to_h.weight": "pytorch_model-02-of-20.bin", "transformer.h.4.mlp.dense_4h_to_h.bias": "pytorch_model-02-of-20.bin", "transformer.h.5.input_layernorm.weight": "pytorch_model-02-of-20.bin", "transformer.h.5.input_layernorm.bias": "pytorch_model-02-of-20.bin", "transformer.h.5.self_attention.query_key_value.weight": "pytorch_model-02-of-20.bin", "transformer.h.5.self_attention.query_key_value.bias": "pytorch_model-02-of-20.bin", "transformer.h.5.self_attention.dense.weight": "pytorch_model-02-of-20.bin", "transformer.h.5.self_attention.dense.bias": "pytorch_model-02-of-20.bin", "transformer.h.5.post_attention_layernorm.weight": "pytorch_model-02-of-20.bin", "transformer.h.5.post_attention_layernorm.bias": "pytorch_model-02-of-20.bin", "transformer.h.5.mlp.dense_h_to_4h.weight": "pytorch_model-02-of-20.bin", "transformer.h.5.mlp.dense_h_to_4h.bias": "pytorch_model-02-of-20.bin", "transformer.h.5.mlp.dense_4h_to_h.weight": "pytorch_model-03-of-20.bin", "transformer.h.5.mlp.dense_4h_to_h.bias": "pytorch_model-03-of-20.bin", "transformer.h.6.input_layernorm.weight": "pytorch_model-03-of-20.bin", "transformer.h.6.input_layernorm.bias": "pytorch_model-03-of-20.bin", "transformer.h.6.self_attention.query_key_value.weight": "pytorch_model-03-of-20.bin", "transformer.h.6.self_attention.query_key_value.bias": "pytorch_model-03-of-20.bin", "transformer.h.6.self_attention.dense.weight": "pytorch_model-03-of-20.bin", "transformer.h.6.self_attention.dense.bias": "pytorch_model-03-of-20.bin", "transformer.h.6.post_attention_layernorm.weight": "pytorch_model-03-of-20.bin", "transformer.h.6.post_attention_layernorm.bias": "pytorch_model-03-of-20.bin", "transformer.h.6.mlp.dense_h_to_4h.weight": "pytorch_model-03-of-20.bin", "transformer.h.6.mlp.dense_h_to_4h.bias": "pytorch_model-03-of-20.bin", "transformer.h.6.mlp.dense_4h_to_h.weight": "pytorch_model-03-of-20.bin", "transformer.h.6.mlp.dense_4h_to_h.bias": "pytorch_model-03-of-20.bin", "transformer.h.7.input_layernorm.weight": "pytorch_model-03-of-20.bin", "transformer.h.7.input_layernorm.bias": "pytorch_model-03-of-20.bin", "transformer.h.7.self_attention.query_key_value.weight": "pytorch_model-03-of-20.bin", "transformer.h.7.self_attention.query_key_value.bias": "pytorch_model-03-of-20.bin", "transformer.h.7.self_attention.dense.weight": "pytorch_model-03-of-20.bin", "transformer.h.7.self_attention.dense.bias": "pytorch_model-03-of-20.bin", "transformer.h.7.post_attention_layernorm.weight": "pytorch_model-03-of-20.bin", "transformer.h.7.post_attention_layernorm.bias": "pytorch_model-03-of-20.bin", "transformer.h.7.mlp.dense_h_to_4h.weight": "pytorch_model-03-of-20.bin", "transformer.h.7.mlp.dense_h_to_4h.bias": "pytorch_model-03-of-20.bin", "transformer.h.7.mlp.dense_4h_to_h.weight": "pytorch_model-03-of-20.bin", "transformer.h.7.mlp.dense_4h_to_h.bias": "pytorch_model-03-of-20.bin", "transformer.h.8.input_layernorm.weight": "pytorch_model-03-of-20.bin", "transformer.h.8.input_layernorm.bias": "pytorch_model-03-of-20.bin", "transformer.h.8.self_attention.query_key_value.weight": "pytorch_model-03-of-20.bin", "transformer.h.8.self_attention.query_key_value.bias": "pytorch_model-03-of-20.bin", "transformer.h.8.self_attention.dense.weight": "pytorch_model-03-of-20.bin", "transformer.h.8.self_attention.dense.bias": "pytorch_model-03-of-20.bin", "transformer.h.8.post_attention_layernorm.weight": "pytorch_model-03-of-20.bin", "transformer.h.8.post_attention_layernorm.bias": "pytorch_model-03-of-20.bin", "transformer.h.8.mlp.dense_h_to_4h.weight": "pytorch_model-03-of-20.bin", "transformer.h.8.mlp.dense_h_to_4h.bias": "pytorch_model-03-of-20.bin", "transformer.h.8.mlp.dense_4h_to_h.weight": "pytorch_model-03-of-20.bin", "transformer.h.8.mlp.dense_4h_to_h.bias": "pytorch_model-03-of-20.bin", "transformer.h.9.input_layernorm.weight": "pytorch_model-03-of-20.bin", "transformer.h.9.input_layernorm.bias": "pytorch_model-03-of-20.bin", "transformer.h.9.self_attention.query_key_value.weight": "pytorch_model-03-of-20.bin", "transformer.h.9.self_attention.query_key_value.bias": "pytorch_model-03-of-20.bin", "transformer.h.9.self_attention.dense.weight": "pytorch_model-03-of-20.bin", "transformer.h.9.self_attention.dense.bias": "pytorch_model-03-of-20.bin", "transformer.h.9.post_attention_layernorm.weight": "pytorch_model-03-of-20.bin", "transformer.h.9.post_attention_layernorm.bias": "pytorch_model-03-of-20.bin", "transformer.h.9.mlp.dense_h_to_4h.weight": "pytorch_model-04-of-20.bin", "transformer.h.9.mlp.dense_h_to_4h.bias": "pytorch_model-04-of-20.bin", "transformer.h.9.mlp.dense_4h_to_h.weight": "pytorch_model-04-of-20.bin", "transformer.h.9.mlp.dense_4h_to_h.bias": "pytorch_model-04-of-20.bin", "transformer.h.10.input_layernorm.weight": "pytorch_model-04-of-20.bin", "transformer.h.10.input_layernorm.bias": "pytorch_model-04-of-20.bin", "transformer.h.10.self_attention.query_key_value.weight": "pytorch_model-04-of-20.bin", "transformer.h.10.self_attention.query_key_value.bias": "pytorch_model-04-of-20.bin", "transformer.h.10.self_attention.dense.weight": "pytorch_model-04-of-20.bin", "transformer.h.10.self_attention.dense.bias": "pytorch_model-04-of-20.bin", "transformer.h.10.post_attention_layernorm.weight": "pytorch_model-04-of-20.bin", "transformer.h.10.post_attention_layernorm.bias": "pytorch_model-04-of-20.bin", "transformer.h.10.mlp.dense_h_to_4h.weight": "pytorch_model-04-of-20.bin", "transformer.h.10.mlp.dense_h_to_4h.bias": "pytorch_model-04-of-20.bin", "transformer.h.10.mlp.dense_4h_to_h.weight": "pytorch_model-04-of-20.bin", "transformer.h.10.mlp.dense_4h_to_h.bias": "pytorch_model-04-of-20.bin", "transformer.h.11.input_layernorm.weight": "pytorch_model-04-of-20.bin", "transformer.h.11.input_layernorm.bias": "pytorch_model-04-of-20.bin", "transformer.h.11.self_attention.query_key_value.weight": "pytorch_model-04-of-20.bin", "transformer.h.11.self_attention.query_key_value.bias": "pytorch_model-04-of-20.bin", "transformer.h.11.self_attention.dense.weight": "pytorch_model-04-of-20.bin", "transformer.h.11.self_attention.dense.bias": "pytorch_model-04-of-20.bin", "transformer.h.11.post_attention_layernorm.weight": "pytorch_model-04-of-20.bin", "transformer.h.11.post_attention_layernorm.bias": "pytorch_model-04-of-20.bin", "transformer.h.11.mlp.dense_h_to_4h.weight": "pytorch_model-04-of-20.bin", "transformer.h.11.mlp.dense_h_to_4h.bias": "pytorch_model-04-of-20.bin", "transformer.h.11.mlp.dense_4h_to_h.weight": "pytorch_model-04-of-20.bin", "transformer.h.11.mlp.dense_4h_to_h.bias": "pytorch_model-04-of-20.bin", "transformer.h.12.input_layernorm.weight": "pytorch_model-04-of-20.bin", "transformer.h.12.input_layernorm.bias": "pytorch_model-04-of-20.bin", "transformer.h.12.self_attention.query_key_value.weight": "pytorch_model-04-of-20.bin", "transformer.h.12.self_attention.query_key_value.bias": "pytorch_model-04-of-20.bin", "transformer.h.12.self_attention.dense.weight": "pytorch_model-04-of-20.bin", "transformer.h.12.self_attention.dense.bias": "pytorch_model-04-of-20.bin", "transformer.h.12.post_attention_layernorm.weight": "pytorch_model-04-of-20.bin", "transformer.h.12.post_attention_layernorm.bias": "pytorch_model-04-of-20.bin", "transformer.h.12.mlp.dense_h_to_4h.weight": "pytorch_model-04-of-20.bin", "transformer.h.12.mlp.dense_h_to_4h.bias": "pytorch_model-04-of-20.bin", "transformer.h.12.mlp.dense_4h_to_h.weight": "pytorch_model-04-of-20.bin", "transformer.h.12.mlp.dense_4h_to_h.bias": "pytorch_model-04-of-20.bin", "transformer.h.13.input_layernorm.weight": "pytorch_model-04-of-20.bin", "transformer.h.13.input_layernorm.bias": "pytorch_model-04-of-20.bin", "transformer.h.13.self_attention.query_key_value.weight": "pytorch_model-05-of-20.bin", "transformer.h.13.self_attention.query_key_value.bias": "pytorch_model-05-of-20.bin", "transformer.h.13.self_attention.dense.weight": "pytorch_model-05-of-20.bin", "transformer.h.13.self_attention.dense.bias": "pytorch_model-05-of-20.bin", "transformer.h.13.post_attention_layernorm.weight": "pytorch_model-05-of-20.bin", "transformer.h.13.post_attention_layernorm.bias": "pytorch_model-05-of-20.bin", "transformer.h.13.mlp.dense_h_to_4h.weight": "pytorch_model-05-of-20.bin", "transformer.h.13.mlp.dense_h_to_4h.bias": "pytorch_model-05-of-20.bin", "transformer.h.13.mlp.dense_4h_to_h.weight": "pytorch_model-05-of-20.bin", "transformer.h.13.mlp.dense_4h_to_h.bias": "pytorch_model-05-of-20.bin", "transformer.h.14.input_layernorm.weight": "pytorch_model-05-of-20.bin", "transformer.h.14.input_layernorm.bias": "pytorch_model-05-of-20.bin", "transformer.h.14.self_attention.query_key_value.weight": "pytorch_model-05-of-20.bin", "transformer.h.14.self_attention.query_key_value.bias": "pytorch_model-05-of-20.bin", "transformer.h.14.self_attention.dense.weight": "pytorch_model-05-of-20.bin", "transformer.h.14.self_attention.dense.bias": "pytorch_model-05-of-20.bin", "transformer.h.14.post_attention_layernorm.weight": "pytorch_model-05-of-20.bin", "transformer.h.14.post_attention_layernorm.bias": "pytorch_model-05-of-20.bin", "transformer.h.14.mlp.dense_h_to_4h.weight": "pytorch_model-05-of-20.bin", "transformer.h.14.mlp.dense_h_to_4h.bias": "pytorch_model-05-of-20.bin", "transformer.h.14.mlp.dense_4h_to_h.weight": "pytorch_model-05-of-20.bin", "transformer.h.14.mlp.dense_4h_to_h.bias": "pytorch_model-05-of-20.bin", "transformer.h.15.input_layernorm.weight": "pytorch_model-05-of-20.bin", "transformer.h.15.input_layernorm.bias": "pytorch_model-05-of-20.bin", "transformer.h.15.self_attention.query_key_value.weight": "pytorch_model-05-of-20.bin", "transformer.h.15.self_attention.query_key_value.bias": "pytorch_model-05-of-20.bin", "transformer.h.15.self_attention.dense.weight": "pytorch_model-05-of-20.bin", "transformer.h.15.self_attention.dense.bias": "pytorch_model-05-of-20.bin", "transformer.h.15.post_attention_layernorm.weight": "pytorch_model-05-of-20.bin", "transformer.h.15.post_attention_layernorm.bias": "pytorch_model-05-of-20.bin", "transformer.h.15.mlp.dense_h_to_4h.weight": "pytorch_model-05-of-20.bin", "transformer.h.15.mlp.dense_h_to_4h.bias": "pytorch_model-05-of-20.bin", "transformer.h.15.mlp.dense_4h_to_h.weight": "pytorch_model-05-of-20.bin", "transformer.h.15.mlp.dense_4h_to_h.bias": "pytorch_model-05-of-20.bin", "transformer.h.16.input_layernorm.weight": "pytorch_model-05-of-20.bin", "transformer.h.16.input_layernorm.bias": "pytorch_model-05-of-20.bin", "transformer.h.16.self_attention.query_key_value.weight": "pytorch_model-05-of-20.bin", "transformer.h.16.self_attention.query_key_value.bias": "pytorch_model-05-of-20.bin", "transformer.h.16.self_attention.dense.weight": "pytorch_model-05-of-20.bin", "transformer.h.16.self_attention.dense.bias": "pytorch_model-05-of-20.bin", "transformer.h.16.post_attention_layernorm.weight": "pytorch_model-05-of-20.bin", "transformer.h.16.post_attention_layernorm.bias": "pytorch_model-05-of-20.bin", "transformer.h.16.mlp.dense_h_to_4h.weight": "pytorch_model-05-of-20.bin", "transformer.h.16.mlp.dense_h_to_4h.bias": "pytorch_model-05-of-20.bin", "transformer.h.16.mlp.dense_4h_to_h.weight": "pytorch_model-06-of-20.bin", "transformer.h.16.mlp.dense_4h_to_h.bias": "pytorch_model-06-of-20.bin", "transformer.h.17.input_layernorm.weight": "pytorch_model-06-of-20.bin", "transformer.h.17.input_layernorm.bias": "pytorch_model-06-of-20.bin", "transformer.h.17.self_attention.query_key_value.weight": "pytorch_model-06-of-20.bin", "transformer.h.17.self_attention.query_key_value.bias": "pytorch_model-06-of-20.bin", "transformer.h.17.self_attention.dense.weight": "pytorch_model-06-of-20.bin", "transformer.h.17.self_attention.dense.bias": "pytorch_model-06-of-20.bin", "transformer.h.17.post_attention_layernorm.weight": "pytorch_model-06-of-20.bin", "transformer.h.17.post_attention_layernorm.bias": "pytorch_model-06-of-20.bin", "transformer.h.17.mlp.dense_h_to_4h.weight": "pytorch_model-06-of-20.bin", "transformer.h.17.mlp.dense_h_to_4h.bias": "pytorch_model-06-of-20.bin", "transformer.h.17.mlp.dense_4h_to_h.weight": "pytorch_model-06-of-20.bin", "transformer.h.17.mlp.dense_4h_to_h.bias": "pytorch_model-06-of-20.bin", "transformer.h.18.input_layernorm.weight": "pytorch_model-06-of-20.bin", "transformer.h.18.input_layernorm.bias": "pytorch_model-06-of-20.bin", "transformer.h.18.self_attention.query_key_value.weight": "pytorch_model-06-of-20.bin", "transformer.h.18.self_attention.query_key_value.bias": "pytorch_model-06-of-20.bin", "transformer.h.18.self_attention.dense.weight": "pytorch_model-06-of-20.bin", "transformer.h.18.self_attention.dense.bias": "pytorch_model-06-of-20.bin", "transformer.h.18.post_attention_layernorm.weight": "pytorch_model-06-of-20.bin", "transformer.h.18.post_attention_layernorm.bias": "pytorch_model-06-of-20.bin", "transformer.h.18.mlp.dense_h_to_4h.weight": "pytorch_model-06-of-20.bin", "transformer.h.18.mlp.dense_h_to_4h.bias": "pytorch_model-06-of-20.bin", "transformer.h.18.mlp.dense_4h_to_h.weight": "pytorch_model-06-of-20.bin", "transformer.h.18.mlp.dense_4h_to_h.bias": "pytorch_model-06-of-20.bin", "transformer.h.19.input_layernorm.weight": "pytorch_model-06-of-20.bin", "transformer.h.19.input_layernorm.bias": "pytorch_model-06-of-20.bin", "transformer.h.19.self_attention.query_key_value.weight": "pytorch_model-06-of-20.bin", "transformer.h.19.self_attention.query_key_value.bias": "pytorch_model-06-of-20.bin", "transformer.h.19.self_attention.dense.weight": "pytorch_model-06-of-20.bin", "transformer.h.19.self_attention.dense.bias": "pytorch_model-06-of-20.bin", "transformer.h.19.post_attention_layernorm.weight": "pytorch_model-06-of-20.bin", "transformer.h.19.post_attention_layernorm.bias": "pytorch_model-06-of-20.bin", "transformer.h.19.mlp.dense_h_to_4h.weight": "pytorch_model-06-of-20.bin", "transformer.h.19.mlp.dense_h_to_4h.bias": "pytorch_model-06-of-20.bin", "transformer.h.19.mlp.dense_4h_to_h.weight": "pytorch_model-06-of-20.bin", "transformer.h.19.mlp.dense_4h_to_h.bias": "pytorch_model-06-of-20.bin", "transformer.h.20.input_layernorm.weight": "pytorch_model-06-of-20.bin", "transformer.h.20.input_layernorm.bias": "pytorch_model-06-of-20.bin", "transformer.h.20.self_attention.query_key_value.weight": "pytorch_model-06-of-20.bin", "transformer.h.20.self_attention.query_key_value.bias": "pytorch_model-06-of-20.bin", "transformer.h.20.self_attention.dense.weight": "pytorch_model-06-of-20.bin", "transformer.h.20.self_attention.dense.bias": "pytorch_model-06-of-20.bin", "transformer.h.20.post_attention_layernorm.weight": "pytorch_model-06-of-20.bin", "transformer.h.20.post_attention_layernorm.bias": "pytorch_model-06-of-20.bin", "transformer.h.20.mlp.dense_h_to_4h.weight": "pytorch_model-07-of-20.bin", "transformer.h.20.mlp.dense_h_to_4h.bias": "pytorch_model-07-of-20.bin", "transformer.h.20.mlp.dense_4h_to_h.weight": "pytorch_model-07-of-20.bin", "transformer.h.20.mlp.dense_4h_to_h.bias": "pytorch_model-07-of-20.bin", "transformer.h.21.input_layernorm.weight": "pytorch_model-07-of-20.bin", "transformer.h.21.input_layernorm.bias": "pytorch_model-07-of-20.bin", "transformer.h.21.self_attention.query_key_value.weight": "pytorch_model-07-of-20.bin", "transformer.h.21.self_attention.query_key_value.bias": "pytorch_model-07-of-20.bin", "transformer.h.21.self_attention.dense.weight": "pytorch_model-07-of-20.bin", "transformer.h.21.self_attention.dense.bias": "pytorch_model-07-of-20.bin", "transformer.h.21.post_attention_layernorm.weight": "pytorch_model-07-of-20.bin", "transformer.h.21.post_attention_layernorm.bias": "pytorch_model-07-of-20.bin", "transformer.h.21.mlp.dense_h_to_4h.weight": "pytorch_model-07-of-20.bin", "transformer.h.21.mlp.dense_h_to_4h.bias": "pytorch_model-07-of-20.bin", "transformer.h.21.mlp.dense_4h_to_h.weight": "pytorch_model-07-of-20.bin", "transformer.h.21.mlp.dense_4h_to_h.bias": "pytorch_model-07-of-20.bin", "transformer.h.22.input_layernorm.weight": "pytorch_model-07-of-20.bin", "transformer.h.22.input_layernorm.bias": "pytorch_model-07-of-20.bin", "transformer.h.22.self_attention.query_key_value.weight": "pytorch_model-07-of-20.bin", "transformer.h.22.self_attention.query_key_value.bias": "pytorch_model-07-of-20.bin", "transformer.h.22.self_attention.dense.weight": "pytorch_model-07-of-20.bin", "transformer.h.22.self_attention.dense.bias": "pytorch_model-07-of-20.bin", "transformer.h.22.post_attention_layernorm.weight": "pytorch_model-07-of-20.bin", "transformer.h.22.post_attention_layernorm.bias": "pytorch_model-07-of-20.bin", "transformer.h.22.mlp.dense_h_to_4h.weight": "pytorch_model-07-of-20.bin", "transformer.h.22.mlp.dense_h_to_4h.bias": "pytorch_model-07-of-20.bin", "transformer.h.22.mlp.dense_4h_to_h.weight": "pytorch_model-07-of-20.bin", "transformer.h.22.mlp.dense_4h_to_h.bias": "pytorch_model-07-of-20.bin", "transformer.h.23.input_layernorm.weight": "pytorch_model-07-of-20.bin", "transformer.h.23.input_layernorm.bias": "pytorch_model-07-of-20.bin", "transformer.h.23.self_attention.query_key_value.weight": "pytorch_model-07-of-20.bin", "transformer.h.23.self_attention.query_key_value.bias": "pytorch_model-07-of-20.bin", "transformer.h.23.self_attention.dense.weight": "pytorch_model-07-of-20.bin", "transformer.h.23.self_attention.dense.bias": "pytorch_model-07-of-20.bin", "transformer.h.23.post_attention_layernorm.weight": "pytorch_model-07-of-20.bin", "transformer.h.23.post_attention_layernorm.bias": "pytorch_model-07-of-20.bin", "transformer.h.23.mlp.dense_h_to_4h.weight": "pytorch_model-07-of-20.bin", "transformer.h.23.mlp.dense_h_to_4h.bias": "pytorch_model-07-of-20.bin", "transformer.h.23.mlp.dense_4h_to_h.weight": "pytorch_model-07-of-20.bin", "transformer.h.23.mlp.dense_4h_to_h.bias": "pytorch_model-07-of-20.bin", "transformer.h.24.input_layernorm.weight": "pytorch_model-07-of-20.bin", "transformer.h.24.input_layernorm.bias": "pytorch_model-07-of-20.bin", "transformer.h.24.self_attention.query_key_value.weight": "pytorch_model-08-of-20.bin", "transformer.h.24.self_attention.query_key_value.bias": "pytorch_model-08-of-20.bin", "transformer.h.24.self_attention.dense.weight": "pytorch_model-08-of-20.bin", "transformer.h.24.self_attention.dense.bias": "pytorch_model-08-of-20.bin", "transformer.h.24.post_attention_layernorm.weight": "pytorch_model-08-of-20.bin", "transformer.h.24.post_attention_layernorm.bias": "pytorch_model-08-of-20.bin", "transformer.h.24.mlp.dense_h_to_4h.weight": "pytorch_model-08-of-20.bin", "transformer.h.24.mlp.dense_h_to_4h.bias": "pytorch_model-08-of-20.bin", "transformer.h.24.mlp.dense_4h_to_h.weight": "pytorch_model-08-of-20.bin", "transformer.h.24.mlp.dense_4h_to_h.bias": "pytorch_model-08-of-20.bin", "transformer.h.25.input_layernorm.weight": "pytorch_model-08-of-20.bin", "transformer.h.25.input_layernorm.bias": "pytorch_model-08-of-20.bin", "transformer.h.25.self_attention.query_key_value.weight": "pytorch_model-08-of-20.bin", "transformer.h.25.self_attention.query_key_value.bias": "pytorch_model-08-of-20.bin", "transformer.h.25.self_attention.dense.weight": "pytorch_model-08-of-20.bin", "transformer.h.25.self_attention.dense.bias": "pytorch_model-08-of-20.bin", "transformer.h.25.post_attention_layernorm.weight": "pytorch_model-08-of-20.bin", "transformer.h.25.post_attention_layernorm.bias": "pytorch_model-08-of-20.bin", "transformer.h.25.mlp.dense_h_to_4h.weight": "pytorch_model-08-of-20.bin", "transformer.h.25.mlp.dense_h_to_4h.bias": "pytorch_model-08-of-20.bin", "transformer.h.25.mlp.dense_4h_to_h.weight": "pytorch_model-08-of-20.bin", "transformer.h.25.mlp.dense_4h_to_h.bias": "pytorch_model-08-of-20.bin", "transformer.h.26.input_layernorm.weight": "pytorch_model-08-of-20.bin", "transformer.h.26.input_layernorm.bias": "pytorch_model-08-of-20.bin", "transformer.h.26.self_attention.query_key_value.weight": "pytorch_model-08-of-20.bin", "transformer.h.26.self_attention.query_key_value.bias": "pytorch_model-08-of-20.bin", "transformer.h.26.self_attention.dense.weight": "pytorch_model-08-of-20.bin", "transformer.h.26.self_attention.dense.bias": "pytorch_model-08-of-20.bin", "transformer.h.26.post_attention_layernorm.weight": "pytorch_model-08-of-20.bin", "transformer.h.26.post_attention_layernorm.bias": "pytorch_model-08-of-20.bin", "transformer.h.26.mlp.dense_h_to_4h.weight": "pytorch_model-08-of-20.bin", "transformer.h.26.mlp.dense_h_to_4h.bias": "pytorch_model-08-of-20.bin", "transformer.h.26.mlp.dense_4h_to_h.weight": "pytorch_model-08-of-20.bin", "transformer.h.26.mlp.dense_4h_to_h.bias": "pytorch_model-08-of-20.bin", "transformer.h.27.input_layernorm.weight": "pytorch_model-08-of-20.bin", "transformer.h.27.input_layernorm.bias": "pytorch_model-08-of-20.bin", "transformer.h.27.self_attention.query_key_value.weight": "pytorch_model-08-of-20.bin", "transformer.h.27.self_attention.query_key_value.bias": "pytorch_model-08-of-20.bin", "transformer.h.27.self_attention.dense.weight": "pytorch_model-08-of-20.bin", "transformer.h.27.self_attention.dense.bias": "pytorch_model-08-of-20.bin", "transformer.h.27.post_attention_layernorm.weight": "pytorch_model-08-of-20.bin", "transformer.h.27.post_attention_layernorm.bias": "pytorch_model-08-of-20.bin", "transformer.h.27.mlp.dense_h_to_4h.weight": "pytorch_model-08-of-20.bin", "transformer.h.27.mlp.dense_h_to_4h.bias": "pytorch_model-08-of-20.bin", "transformer.h.27.mlp.dense_4h_to_h.weight": "pytorch_model-09-of-20.bin", "transformer.h.27.mlp.dense_4h_to_h.bias": "pytorch_model-09-of-20.bin", "transformer.h.28.input_layernorm.weight": "pytorch_model-09-of-20.bin", "transformer.h.28.input_layernorm.bias": "pytorch_model-09-of-20.bin", "transformer.h.28.self_attention.query_key_value.weight": "pytorch_model-09-of-20.bin", "transformer.h.28.self_attention.query_key_value.bias": "pytorch_model-09-of-20.bin", "transformer.h.28.self_attention.dense.weight": "pytorch_model-09-of-20.bin", "transformer.h.28.self_attention.dense.bias": "pytorch_model-09-of-20.bin", "transformer.h.28.post_attention_layernorm.weight": "pytorch_model-09-of-20.bin", "transformer.h.28.post_attention_layernorm.bias": "pytorch_model-09-of-20.bin", "transformer.h.28.mlp.dense_h_to_4h.weight": "pytorch_model-09-of-20.bin", "transformer.h.28.mlp.dense_h_to_4h.bias": "pytorch_model-09-of-20.bin", "transformer.h.28.mlp.dense_4h_to_h.weight": "pytorch_model-09-of-20.bin", "transformer.h.28.mlp.dense_4h_to_h.bias": "pytorch_model-09-of-20.bin", "transformer.h.29.input_layernorm.weight": "pytorch_model-09-of-20.bin", "transformer.h.29.input_layernorm.bias": "pytorch_model-09-of-20.bin", "transformer.h.29.self_attention.query_key_value.weight": "pytorch_model-09-of-20.bin", "transformer.h.29.self_attention.query_key_value.bias": "pytorch_model-09-of-20.bin", "transformer.h.29.self_attention.dense.weight": "pytorch_model-09-of-20.bin", "transformer.h.29.self_attention.dense.bias": "pytorch_model-09-of-20.bin", "transformer.h.29.post_attention_layernorm.weight": "pytorch_model-09-of-20.bin", "transformer.h.29.post_attention_layernorm.bias": "pytorch_model-09-of-20.bin", "transformer.h.29.mlp.dense_h_to_4h.weight": "pytorch_model-09-of-20.bin", "transformer.h.29.mlp.dense_h_to_4h.bias": "pytorch_model-09-of-20.bin", "transformer.h.29.mlp.dense_4h_to_h.weight": "pytorch_model-09-of-20.bin", "transformer.h.29.mlp.dense_4h_to_h.bias": "pytorch_model-09-of-20.bin", "transformer.h.30.input_layernorm.weight": "pytorch_model-09-of-20.bin", "transformer.h.30.input_layernorm.bias": "pytorch_model-09-of-20.bin", "transformer.h.30.self_attention.query_key_value.weight": "pytorch_model-09-of-20.bin", "transformer.h.30.self_attention.query_key_value.bias": "pytorch_model-09-of-20.bin", "transformer.h.30.self_attention.dense.weight": "pytorch_model-09-of-20.bin", "transformer.h.30.self_attention.dense.bias": "pytorch_model-09-of-20.bin", "transformer.h.30.post_attention_layernorm.weight": "pytorch_model-09-of-20.bin", "transformer.h.30.post_attention_layernorm.bias": "pytorch_model-09-of-20.bin", "transformer.h.30.mlp.dense_h_to_4h.weight": "pytorch_model-09-of-20.bin", "transformer.h.30.mlp.dense_h_to_4h.bias": "pytorch_model-09-of-20.bin", "transformer.h.30.mlp.dense_4h_to_h.weight": "pytorch_model-09-of-20.bin", "transformer.h.30.mlp.dense_4h_to_h.bias": "pytorch_model-09-of-20.bin", "transformer.h.31.input_layernorm.weight": "pytorch_model-09-of-20.bin", "transformer.h.31.input_layernorm.bias": "pytorch_model-09-of-20.bin", "transformer.h.31.self_attention.query_key_value.weight": "pytorch_model-09-of-20.bin", "transformer.h.31.self_attention.query_key_value.bias": "pytorch_model-09-of-20.bin", "transformer.h.31.self_attention.dense.weight": "pytorch_model-09-of-20.bin", "transformer.h.31.self_attention.dense.bias": "pytorch_model-09-of-20.bin", "transformer.h.31.post_attention_layernorm.weight": "pytorch_model-09-of-20.bin", "transformer.h.31.post_attention_layernorm.bias": "pytorch_model-09-of-20.bin", "transformer.h.31.mlp.dense_h_to_4h.weight": "pytorch_model-10-of-20.bin", "transformer.h.31.mlp.dense_h_to_4h.bias": "pytorch_model-10-of-20.bin", "transformer.h.31.mlp.dense_4h_to_h.weight": "pytorch_model-10-of-20.bin", "transformer.h.31.mlp.dense_4h_to_h.bias": "pytorch_model-10-of-20.bin", "transformer.h.32.input_layernorm.weight": "pytorch_model-10-of-20.bin", "transformer.h.32.input_layernorm.bias": "pytorch_model-10-of-20.bin", "transformer.h.32.self_attention.query_key_value.weight": "pytorch_model-10-of-20.bin", "transformer.h.32.self_attention.query_key_value.bias": "pytorch_model-10-of-20.bin", "transformer.h.32.self_attention.dense.weight": "pytorch_model-10-of-20.bin", "transformer.h.32.self_attention.dense.bias": "pytorch_model-10-of-20.bin", "transformer.h.32.post_attention_layernorm.weight": "pytorch_model-10-of-20.bin", "transformer.h.32.post_attention_layernorm.bias": "pytorch_model-10-of-20.bin", "transformer.h.32.mlp.dense_h_to_4h.weight": "pytorch_model-10-of-20.bin", "transformer.h.32.mlp.dense_h_to_4h.bias": "pytorch_model-10-of-20.bin", "transformer.h.32.mlp.dense_4h_to_h.weight": "pytorch_model-10-of-20.bin", "transformer.h.32.mlp.dense_4h_to_h.bias": "pytorch_model-10-of-20.bin", "transformer.h.33.input_layernorm.weight": "pytorch_model-10-of-20.bin", "transformer.h.33.input_layernorm.bias": "pytorch_model-10-of-20.bin", "transformer.h.33.self_attention.query_key_value.weight": "pytorch_model-10-of-20.bin", "transformer.h.33.self_attention.query_key_value.bias": "pytorch_model-10-of-20.bin", "transformer.h.33.self_attention.dense.weight": "pytorch_model-10-of-20.bin", "transformer.h.33.self_attention.dense.bias": "pytorch_model-10-of-20.bin", "transformer.h.33.post_attention_layernorm.weight": "pytorch_model-10-of-20.bin", "transformer.h.33.post_attention_layernorm.bias": "pytorch_model-10-of-20.bin", "transformer.h.33.mlp.dense_h_to_4h.weight": "pytorch_model-10-of-20.bin", "transformer.h.33.mlp.dense_h_to_4h.bias": "pytorch_model-10-of-20.bin", "transformer.h.33.mlp.dense_4h_to_h.weight": "pytorch_model-10-of-20.bin", "transformer.h.33.mlp.dense_4h_to_h.bias": "pytorch_model-10-of-20.bin", "transformer.h.34.input_layernorm.weight": "pytorch_model-10-of-20.bin", "transformer.h.34.input_layernorm.bias": "pytorch_model-10-of-20.bin", "transformer.h.34.self_attention.query_key_value.weight": "pytorch_model-10-of-20.bin", "transformer.h.34.self_attention.query_key_value.bias": "pytorch_model-10-of-20.bin", "transformer.h.34.self_attention.dense.weight": "pytorch_model-10-of-20.bin", "transformer.h.34.self_attention.dense.bias": "pytorch_model-10-of-20.bin", "transformer.h.34.post_attention_layernorm.weight": "pytorch_model-10-of-20.bin", "transformer.h.34.post_attention_layernorm.bias": "pytorch_model-10-of-20.bin", "transformer.h.34.mlp.dense_h_to_4h.weight": "pytorch_model-10-of-20.bin", "transformer.h.34.mlp.dense_h_to_4h.bias": "pytorch_model-10-of-20.bin", "transformer.h.34.mlp.dense_4h_to_h.weight": "pytorch_model-10-of-20.bin", "transformer.h.34.mlp.dense_4h_to_h.bias": "pytorch_model-10-of-20.bin", "transformer.h.35.input_layernorm.weight": "pytorch_model-10-of-20.bin", "transformer.h.35.input_layernorm.bias": "pytorch_model-10-of-20.bin", "transformer.h.35.self_attention.query_key_value.weight": "pytorch_model-11-of-20.bin", "transformer.h.35.self_attention.query_key_value.bias": "pytorch_model-11-of-20.bin", "transformer.h.35.self_attention.dense.weight": "pytorch_model-11-of-20.bin", "transformer.h.35.self_attention.dense.bias": "pytorch_model-11-of-20.bin", "transformer.h.35.post_attention_layernorm.weight": "pytorch_model-11-of-20.bin", "transformer.h.35.post_attention_layernorm.bias": "pytorch_model-11-of-20.bin", "transformer.h.35.mlp.dense_h_to_4h.weight": "pytorch_model-11-of-20.bin", "transformer.h.35.mlp.dense_h_to_4h.bias": "pytorch_model-11-of-20.bin", "transformer.h.35.mlp.dense_4h_to_h.weight": "pytorch_model-11-of-20.bin", "transformer.h.35.mlp.dense_4h_to_h.bias": "pytorch_model-11-of-20.bin", "transformer.h.36.input_layernorm.weight": "pytorch_model-11-of-20.bin", "transformer.h.36.input_layernorm.bias": "pytorch_model-11-of-20.bin", "transformer.h.36.self_attention.query_key_value.weight": "pytorch_model-11-of-20.bin", "transformer.h.36.self_attention.query_key_value.bias": "pytorch_model-11-of-20.bin", "transformer.h.36.self_attention.dense.weight": "pytorch_model-11-of-20.bin", "transformer.h.36.self_attention.dense.bias": "pytorch_model-11-of-20.bin", "transformer.h.36.post_attention_layernorm.weight": "pytorch_model-11-of-20.bin", "transformer.h.36.post_attention_layernorm.bias": "pytorch_model-11-of-20.bin", "transformer.h.36.mlp.dense_h_to_4h.weight": "pytorch_model-11-of-20.bin", "transformer.h.36.mlp.dense_h_to_4h.bias": "pytorch_model-11-of-20.bin", "transformer.h.36.mlp.dense_4h_to_h.weight": "pytorch_model-11-of-20.bin", "transformer.h.36.mlp.dense_4h_to_h.bias": "pytorch_model-11-of-20.bin", "transformer.h.37.input_layernorm.weight": "pytorch_model-11-of-20.bin", "transformer.h.37.input_layernorm.bias": "pytorch_model-11-of-20.bin", "transformer.h.37.self_attention.query_key_value.weight": "pytorch_model-11-of-20.bin", "transformer.h.37.self_attention.query_key_value.bias": "pytorch_model-11-of-20.bin", "transformer.h.37.self_attention.dense.weight": "pytorch_model-11-of-20.bin", "transformer.h.37.self_attention.dense.bias": "pytorch_model-11-of-20.bin", "transformer.h.37.post_attention_layernorm.weight": "pytorch_model-11-of-20.bin", "transformer.h.37.post_attention_layernorm.bias": "pytorch_model-11-of-20.bin", "transformer.h.37.mlp.dense_h_to_4h.weight": "pytorch_model-11-of-20.bin", "transformer.h.37.mlp.dense_h_to_4h.bias": "pytorch_model-11-of-20.bin", "transformer.h.37.mlp.dense_4h_to_h.weight": "pytorch_model-11-of-20.bin", "transformer.h.37.mlp.dense_4h_to_h.bias": "pytorch_model-11-of-20.bin", "transformer.h.38.input_layernorm.weight": "pytorch_model-11-of-20.bin", "transformer.h.38.input_layernorm.bias": "pytorch_model-11-of-20.bin", "transformer.h.38.self_attention.query_key_value.weight": "pytorch_model-11-of-20.bin", "transformer.h.38.self_attention.query_key_value.bias": "pytorch_model-11-of-20.bin", "transformer.h.38.self_attention.dense.weight": "pytorch_model-11-of-20.bin", "transformer.h.38.self_attention.dense.bias": "pytorch_model-11-of-20.bin", "transformer.h.38.post_attention_layernorm.weight": "pytorch_model-11-of-20.bin", "transformer.h.38.post_attention_layernorm.bias": "pytorch_model-11-of-20.bin", "transformer.h.38.mlp.dense_h_to_4h.weight": "pytorch_model-11-of-20.bin", "transformer.h.38.mlp.dense_h_to_4h.bias": "pytorch_model-11-of-20.bin", "transformer.h.38.mlp.dense_4h_to_h.weight": "pytorch_model-12-of-20.bin", "transformer.h.38.mlp.dense_4h_to_h.bias": "pytorch_model-12-of-20.bin", "transformer.h.39.input_layernorm.weight": "pytorch_model-12-of-20.bin", "transformer.h.39.input_layernorm.bias": "pytorch_model-12-of-20.bin", "transformer.h.39.self_attention.query_key_value.weight": "pytorch_model-12-of-20.bin", "transformer.h.39.self_attention.query_key_value.bias": "pytorch_model-12-of-20.bin", "transformer.h.39.self_attention.dense.weight": "pytorch_model-12-of-20.bin", "transformer.h.39.self_attention.dense.bias": "pytorch_model-12-of-20.bin", "transformer.h.39.post_attention_layernorm.weight": "pytorch_model-12-of-20.bin", "transformer.h.39.post_attention_layernorm.bias": "pytorch_model-12-of-20.bin", "transformer.h.39.mlp.dense_h_to_4h.weight": "pytorch_model-12-of-20.bin", "transformer.h.39.mlp.dense_h_to_4h.bias": "pytorch_model-12-of-20.bin", "transformer.h.39.mlp.dense_4h_to_h.weight": "pytorch_model-12-of-20.bin", "transformer.h.39.mlp.dense_4h_to_h.bias": "pytorch_model-12-of-20.bin", "transformer.h.40.input_layernorm.weight": "pytorch_model-12-of-20.bin", "transformer.h.40.input_layernorm.bias": "pytorch_model-12-of-20.bin", "transformer.h.40.self_attention.query_key_value.weight": "pytorch_model-12-of-20.bin", "transformer.h.40.self_attention.query_key_value.bias": "pytorch_model-12-of-20.bin", "transformer.h.40.self_attention.dense.weight": "pytorch_model-12-of-20.bin", "transformer.h.40.self_attention.dense.bias": "pytorch_model-12-of-20.bin", "transformer.h.40.post_attention_layernorm.weight": "pytorch_model-12-of-20.bin", "transformer.h.40.post_attention_layernorm.bias": "pytorch_model-12-of-20.bin", "transformer.h.40.mlp.dense_h_to_4h.weight": "pytorch_model-12-of-20.bin", "transformer.h.40.mlp.dense_h_to_4h.bias": "pytorch_model-12-of-20.bin", "transformer.h.40.mlp.dense_4h_to_h.weight": "pytorch_model-12-of-20.bin", "transformer.h.40.mlp.dense_4h_to_h.bias": "pytorch_model-12-of-20.bin", "transformer.h.41.input_layernorm.weight": "pytorch_model-12-of-20.bin", "transformer.h.41.input_layernorm.bias": "pytorch_model-12-of-20.bin", "transformer.h.41.self_attention.query_key_value.weight": "pytorch_model-12-of-20.bin", "transformer.h.41.self_attention.query_key_value.bias": "pytorch_model-12-of-20.bin", "transformer.h.41.self_attention.dense.weight": "pytorch_model-12-of-20.bin", "transformer.h.41.self_attention.dense.bias": "pytorch_model-12-of-20.bin", "transformer.h.41.post_attention_layernorm.weight": "pytorch_model-12-of-20.bin", "transformer.h.41.post_attention_layernorm.bias": "pytorch_model-12-of-20.bin", "transformer.h.41.mlp.dense_h_to_4h.weight": "pytorch_model-12-of-20.bin", "transformer.h.41.mlp.dense_h_to_4h.bias": "pytorch_model-12-of-20.bin", "transformer.h.41.mlp.dense_4h_to_h.weight": "pytorch_model-12-of-20.bin", "transformer.h.41.mlp.dense_4h_to_h.bias": "pytorch_model-12-of-20.bin", "transformer.h.42.input_layernorm.weight": "pytorch_model-12-of-20.bin", "transformer.h.42.input_layernorm.bias": "pytorch_model-12-of-20.bin", "transformer.h.42.self_attention.query_key_value.weight": "pytorch_model-12-of-20.bin", "transformer.h.42.self_attention.query_key_value.bias": "pytorch_model-12-of-20.bin", "transformer.h.42.self_attention.dense.weight": "pytorch_model-12-of-20.bin", "transformer.h.42.self_attention.dense.bias": "pytorch_model-12-of-20.bin", "transformer.h.42.post_attention_layernorm.weight": "pytorch_model-12-of-20.bin", "transformer.h.42.post_attention_layernorm.bias": "pytorch_model-12-of-20.bin", "transformer.h.42.mlp.dense_h_to_4h.weight": "pytorch_model-13-of-20.bin", "transformer.h.42.mlp.dense_h_to_4h.bias": "pytorch_model-13-of-20.bin", "transformer.h.42.mlp.dense_4h_to_h.weight": "pytorch_model-13-of-20.bin", "transformer.h.42.mlp.dense_4h_to_h.bias": "pytorch_model-13-of-20.bin", "transformer.h.43.input_layernorm.weight": "pytorch_model-13-of-20.bin", "transformer.h.43.input_layernorm.bias": "pytorch_model-13-of-20.bin", "transformer.h.43.self_attention.query_key_value.weight": "pytorch_model-13-of-20.bin", "transformer.h.43.self_attention.query_key_value.bias": "pytorch_model-13-of-20.bin", "transformer.h.43.self_attention.dense.weight": "pytorch_model-13-of-20.bin", "transformer.h.43.self_attention.dense.bias": "pytorch_model-13-of-20.bin", "transformer.h.43.post_attention_layernorm.weight": "pytorch_model-13-of-20.bin", "transformer.h.43.post_attention_layernorm.bias": "pytorch_model-13-of-20.bin", "transformer.h.43.mlp.dense_h_to_4h.weight": "pytorch_model-13-of-20.bin", "transformer.h.43.mlp.dense_h_to_4h.bias": "pytorch_model-13-of-20.bin", "transformer.h.43.mlp.dense_4h_to_h.weight": "pytorch_model-13-of-20.bin", "transformer.h.43.mlp.dense_4h_to_h.bias": "pytorch_model-13-of-20.bin", "transformer.h.44.input_layernorm.weight": "pytorch_model-13-of-20.bin", "transformer.h.44.input_layernorm.bias": "pytorch_model-13-of-20.bin", "transformer.h.44.self_attention.query_key_value.weight": "pytorch_model-13-of-20.bin", "transformer.h.44.self_attention.query_key_value.bias": "pytorch_model-13-of-20.bin", "transformer.h.44.self_attention.dense.weight": "pytorch_model-13-of-20.bin", "transformer.h.44.self_attention.dense.bias": "pytorch_model-13-of-20.bin", "transformer.h.44.post_attention_layernorm.weight": "pytorch_model-13-of-20.bin", "transformer.h.44.post_attention_layernorm.bias": "pytorch_model-13-of-20.bin", "transformer.h.44.mlp.dense_h_to_4h.weight": "pytorch_model-13-of-20.bin", "transformer.h.44.mlp.dense_h_to_4h.bias": "pytorch_model-13-of-20.bin", "transformer.h.44.mlp.dense_4h_to_h.weight": "pytorch_model-13-of-20.bin", "transformer.h.44.mlp.dense_4h_to_h.bias": "pytorch_model-13-of-20.bin", "transformer.h.45.input_layernorm.weight": "pytorch_model-13-of-20.bin", "transformer.h.45.input_layernorm.bias": "pytorch_model-13-of-20.bin", "transformer.h.45.self_attention.query_key_value.weight": "pytorch_model-13-of-20.bin", "transformer.h.45.self_attention.query_key_value.bias": "pytorch_model-13-of-20.bin", "transformer.h.45.self_attention.dense.weight": "pytorch_model-13-of-20.bin", "transformer.h.45.self_attention.dense.bias": "pytorch_model-13-of-20.bin", "transformer.h.45.post_attention_layernorm.weight": "pytorch_model-13-of-20.bin", "transformer.h.45.post_attention_layernorm.bias": "pytorch_model-13-of-20.bin", "transformer.h.45.mlp.dense_h_to_4h.weight": "pytorch_model-13-of-20.bin", "transformer.h.45.mlp.dense_h_to_4h.bias": "pytorch_model-13-of-20.bin", "transformer.h.45.mlp.dense_4h_to_h.weight": "pytorch_model-13-of-20.bin", "transformer.h.45.mlp.dense_4h_to_h.bias": "pytorch_model-13-of-20.bin", "transformer.h.46.input_layernorm.weight": "pytorch_model-13-of-20.bin", "transformer.h.46.input_layernorm.bias": "pytorch_model-13-of-20.bin", "transformer.h.46.self_attention.query_key_value.weight": "pytorch_model-14-of-20.bin", "transformer.h.46.self_attention.query_key_value.bias": "pytorch_model-14-of-20.bin", "transformer.h.46.self_attention.dense.weight": "pytorch_model-14-of-20.bin", "transformer.h.46.self_attention.dense.bias": "pytorch_model-14-of-20.bin", "transformer.h.46.post_attention_layernorm.weight": "pytorch_model-14-of-20.bin", "transformer.h.46.post_attention_layernorm.bias": "pytorch_model-14-of-20.bin", "transformer.h.46.mlp.dense_h_to_4h.weight": "pytorch_model-14-of-20.bin", "transformer.h.46.mlp.dense_h_to_4h.bias": "pytorch_model-14-of-20.bin", "transformer.h.46.mlp.dense_4h_to_h.weight": "pytorch_model-14-of-20.bin", "transformer.h.46.mlp.dense_4h_to_h.bias": "pytorch_model-14-of-20.bin", "transformer.h.47.input_layernorm.weight": "pytorch_model-14-of-20.bin", "transformer.h.47.input_layernorm.bias": "pytorch_model-14-of-20.bin", "transformer.h.47.self_attention.query_key_value.weight": "pytorch_model-14-of-20.bin", "transformer.h.47.self_attention.query_key_value.bias": "pytorch_model-14-of-20.bin", "transformer.h.47.self_attention.dense.weight": "pytorch_model-14-of-20.bin", "transformer.h.47.self_attention.dense.bias": "pytorch_model-14-of-20.bin", "transformer.h.47.post_attention_layernorm.weight": "pytorch_model-14-of-20.bin", "transformer.h.47.post_attention_layernorm.bias": "pytorch_model-14-of-20.bin", "transformer.h.47.mlp.dense_h_to_4h.weight": "pytorch_model-14-of-20.bin", "transformer.h.47.mlp.dense_h_to_4h.bias": "pytorch_model-14-of-20.bin", "transformer.h.47.mlp.dense_4h_to_h.weight": "pytorch_model-14-of-20.bin", "transformer.h.47.mlp.dense_4h_to_h.bias": "pytorch_model-14-of-20.bin", "transformer.h.48.input_layernorm.weight": "pytorch_model-14-of-20.bin", "transformer.h.48.input_layernorm.bias": "pytorch_model-14-of-20.bin", "transformer.h.48.self_attention.query_key_value.weight": "pytorch_model-14-of-20.bin", "transformer.h.48.self_attention.query_key_value.bias": "pytorch_model-14-of-20.bin", "transformer.h.48.self_attention.dense.weight": "pytorch_model-14-of-20.bin", "transformer.h.48.self_attention.dense.bias": "pytorch_model-14-of-20.bin", "transformer.h.48.post_attention_layernorm.weight": "pytorch_model-14-of-20.bin", "transformer.h.48.post_attention_layernorm.bias": "pytorch_model-14-of-20.bin", "transformer.h.48.mlp.dense_h_to_4h.weight": "pytorch_model-14-of-20.bin", "transformer.h.48.mlp.dense_h_to_4h.bias": "pytorch_model-14-of-20.bin", "transformer.h.48.mlp.dense_4h_to_h.weight": "pytorch_model-14-of-20.bin", "transformer.h.48.mlp.dense_4h_to_h.bias": "pytorch_model-14-of-20.bin", "transformer.h.49.input_layernorm.weight": "pytorch_model-14-of-20.bin", "transformer.h.49.input_layernorm.bias": "pytorch_model-14-of-20.bin", "transformer.h.49.self_attention.query_key_value.weight": "pytorch_model-14-of-20.bin", "transformer.h.49.self_attention.query_key_value.bias": "pytorch_model-14-of-20.bin", "transformer.h.49.self_attention.dense.weight": "pytorch_model-14-of-20.bin", "transformer.h.49.self_attention.dense.bias": "pytorch_model-14-of-20.bin", "transformer.h.49.post_attention_layernorm.weight": "pytorch_model-14-of-20.bin", "transformer.h.49.post_attention_layernorm.bias": "pytorch_model-14-of-20.bin", "transformer.h.49.mlp.dense_h_to_4h.weight": "pytorch_model-14-of-20.bin", "transformer.h.49.mlp.dense_h_to_4h.bias": "pytorch_model-14-of-20.bin", "transformer.h.49.mlp.dense_4h_to_h.weight": "pytorch_model-15-of-20.bin", "transformer.h.49.mlp.dense_4h_to_h.bias": "pytorch_model-15-of-20.bin", "transformer.h.50.input_layernorm.weight": "pytorch_model-15-of-20.bin", "transformer.h.50.input_layernorm.bias": "pytorch_model-15-of-20.bin", "transformer.h.50.self_attention.query_key_value.weight": "pytorch_model-15-of-20.bin", "transformer.h.50.self_attention.query_key_value.bias": "pytorch_model-15-of-20.bin", "transformer.h.50.self_attention.dense.weight": "pytorch_model-15-of-20.bin", "transformer.h.50.self_attention.dense.bias": "pytorch_model-15-of-20.bin", "transformer.h.50.post_attention_layernorm.weight": "pytorch_model-15-of-20.bin", "transformer.h.50.post_attention_layernorm.bias": "pytorch_model-15-of-20.bin", "transformer.h.50.mlp.dense_h_to_4h.weight": "pytorch_model-15-of-20.bin", "transformer.h.50.mlp.dense_h_to_4h.bias": "pytorch_model-15-of-20.bin", "transformer.h.50.mlp.dense_4h_to_h.weight": "pytorch_model-15-of-20.bin", "transformer.h.50.mlp.dense_4h_to_h.bias": "pytorch_model-15-of-20.bin", "transformer.h.51.input_layernorm.weight": "pytorch_model-15-of-20.bin", "transformer.h.51.input_layernorm.bias": "pytorch_model-15-of-20.bin", "transformer.h.51.self_attention.query_key_value.weight": "pytorch_model-15-of-20.bin", "transformer.h.51.self_attention.query_key_value.bias": "pytorch_model-15-of-20.bin", "transformer.h.51.self_attention.dense.weight": "pytorch_model-15-of-20.bin", "transformer.h.51.self_attention.dense.bias": "pytorch_model-15-of-20.bin", "transformer.h.51.post_attention_layernorm.weight": "pytorch_model-15-of-20.bin", "transformer.h.51.post_attention_layernorm.bias": "pytorch_model-15-of-20.bin", "transformer.h.51.mlp.dense_h_to_4h.weight": "pytorch_model-15-of-20.bin", "transformer.h.51.mlp.dense_h_to_4h.bias": "pytorch_model-15-of-20.bin", "transformer.h.51.mlp.dense_4h_to_h.weight": "pytorch_model-15-of-20.bin", "transformer.h.51.mlp.dense_4h_to_h.bias": "pytorch_model-15-of-20.bin", "transformer.h.52.input_layernorm.weight": "pytorch_model-15-of-20.bin", "transformer.h.52.input_layernorm.bias": "pytorch_model-15-of-20.bin", "transformer.h.52.self_attention.query_key_value.weight": "pytorch_model-15-of-20.bin", "transformer.h.52.self_attention.query_key_value.bias": "pytorch_model-15-of-20.bin", "transformer.h.52.self_attention.dense.weight": "pytorch_model-15-of-20.bin", "transformer.h.52.self_attention.dense.bias": "pytorch_model-15-of-20.bin", "transformer.h.52.post_attention_layernorm.weight": "pytorch_model-15-of-20.bin", "transformer.h.52.post_attention_layernorm.bias": "pytorch_model-15-of-20.bin", "transformer.h.52.mlp.dense_h_to_4h.weight": "pytorch_model-15-of-20.bin", "transformer.h.52.mlp.dense_h_to_4h.bias": "pytorch_model-15-of-20.bin", "transformer.h.52.mlp.dense_4h_to_h.weight": "pytorch_model-15-of-20.bin", "transformer.h.52.mlp.dense_4h_to_h.bias": "pytorch_model-15-of-20.bin", "transformer.h.53.input_layernorm.weight": "pytorch_model-15-of-20.bin", "transformer.h.53.input_layernorm.bias": "pytorch_model-15-of-20.bin", "transformer.h.53.self_attention.query_key_value.weight": "pytorch_model-15-of-20.bin", "transformer.h.53.self_attention.query_key_value.bias": "pytorch_model-15-of-20.bin", "transformer.h.53.self_attention.dense.weight": "pytorch_model-15-of-20.bin", "transformer.h.53.self_attention.dense.bias": "pytorch_model-15-of-20.bin", "transformer.h.53.post_attention_layernorm.weight": "pytorch_model-15-of-20.bin", "transformer.h.53.post_attention_layernorm.bias": "pytorch_model-15-of-20.bin", "transformer.h.53.mlp.dense_h_to_4h.weight": "pytorch_model-16-of-20.bin", "transformer.h.53.mlp.dense_h_to_4h.bias": "pytorch_model-16-of-20.bin", "transformer.h.53.mlp.dense_4h_to_h.weight": "pytorch_model-16-of-20.bin", "transformer.h.53.mlp.dense_4h_to_h.bias": "pytorch_model-16-of-20.bin", "transformer.h.54.input_layernorm.weight": "pytorch_model-16-of-20.bin", "transformer.h.54.input_layernorm.bias": "pytorch_model-16-of-20.bin", "transformer.h.54.self_attention.query_key_value.weight": "pytorch_model-16-of-20.bin", "transformer.h.54.self_attention.query_key_value.bias": "pytorch_model-16-of-20.bin", "transformer.h.54.self_attention.dense.weight": "pytorch_model-16-of-20.bin", "transformer.h.54.self_attention.dense.bias": "pytorch_model-16-of-20.bin", "transformer.h.54.post_attention_layernorm.weight": "pytorch_model-16-of-20.bin", "transformer.h.54.post_attention_layernorm.bias": "pytorch_model-16-of-20.bin", "transformer.h.54.mlp.dense_h_to_4h.weight": "pytorch_model-16-of-20.bin", "transformer.h.54.mlp.dense_h_to_4h.bias": "pytorch_model-16-of-20.bin", "transformer.h.54.mlp.dense_4h_to_h.weight": "pytorch_model-16-of-20.bin", "transformer.h.54.mlp.dense_4h_to_h.bias": "pytorch_model-16-of-20.bin", "transformer.h.55.input_layernorm.weight": "pytorch_model-16-of-20.bin", "transformer.h.55.input_layernorm.bias": "pytorch_model-16-of-20.bin", "transformer.h.55.self_attention.query_key_value.weight": "pytorch_model-16-of-20.bin", "transformer.h.55.self_attention.query_key_value.bias": "pytorch_model-16-of-20.bin", "transformer.h.55.self_attention.dense.weight": "pytorch_model-16-of-20.bin", "transformer.h.55.self_attention.dense.bias": "pytorch_model-16-of-20.bin", "transformer.h.55.post_attention_layernorm.weight": "pytorch_model-16-of-20.bin", "transformer.h.55.post_attention_layernorm.bias": "pytorch_model-16-of-20.bin", "transformer.h.55.mlp.dense_h_to_4h.weight": "pytorch_model-16-of-20.bin", "transformer.h.55.mlp.dense_h_to_4h.bias": "pytorch_model-16-of-20.bin", "transformer.h.55.mlp.dense_4h_to_h.weight": "pytorch_model-16-of-20.bin", "transformer.h.55.mlp.dense_4h_to_h.bias": "pytorch_model-16-of-20.bin", "transformer.h.56.input_layernorm.weight": "pytorch_model-16-of-20.bin", "transformer.h.56.input_layernorm.bias": "pytorch_model-16-of-20.bin", "transformer.h.56.self_attention.query_key_value.weight": "pytorch_model-16-of-20.bin", "transformer.h.56.self_attention.query_key_value.bias": "pytorch_model-16-of-20.bin", "transformer.h.56.self_attention.dense.weight": "pytorch_model-16-of-20.bin", "transformer.h.56.self_attention.dense.bias": "pytorch_model-16-of-20.bin", "transformer.h.56.post_attention_layernorm.weight": "pytorch_model-16-of-20.bin", "transformer.h.56.post_attention_layernorm.bias": "pytorch_model-16-of-20.bin", "transformer.h.56.mlp.dense_h_to_4h.weight": "pytorch_model-16-of-20.bin", "transformer.h.56.mlp.dense_h_to_4h.bias": "pytorch_model-16-of-20.bin", "transformer.h.56.mlp.dense_4h_to_h.weight": "pytorch_model-16-of-20.bin", "transformer.h.56.mlp.dense_4h_to_h.bias": "pytorch_model-16-of-20.bin", "transformer.h.57.input_layernorm.weight": "pytorch_model-16-of-20.bin", "transformer.h.57.input_layernorm.bias": "pytorch_model-16-of-20.bin", "transformer.h.57.self_attention.query_key_value.weight": "pytorch_model-17-of-20.bin", "transformer.h.57.self_attention.query_key_value.bias": "pytorch_model-17-of-20.bin", "transformer.h.57.self_attention.dense.weight": "pytorch_model-17-of-20.bin", "transformer.h.57.self_attention.dense.bias": "pytorch_model-17-of-20.bin", "transformer.h.57.post_attention_layernorm.weight": "pytorch_model-17-of-20.bin", "transformer.h.57.post_attention_layernorm.bias": "pytorch_model-17-of-20.bin", "transformer.h.57.mlp.dense_h_to_4h.weight": "pytorch_model-17-of-20.bin", "transformer.h.57.mlp.dense_h_to_4h.bias": "pytorch_model-17-of-20.bin", "transformer.h.57.mlp.dense_4h_to_h.weight": "pytorch_model-17-of-20.bin", "transformer.h.57.mlp.dense_4h_to_h.bias": "pytorch_model-17-of-20.bin", "transformer.h.58.input_layernorm.weight": "pytorch_model-17-of-20.bin", "transformer.h.58.input_layernorm.bias": "pytorch_model-17-of-20.bin", "transformer.h.58.self_attention.query_key_value.weight": "pytorch_model-17-of-20.bin", "transformer.h.58.self_attention.query_key_value.bias": "pytorch_model-17-of-20.bin", "transformer.h.58.self_attention.dense.weight": "pytorch_model-17-of-20.bin", "transformer.h.58.self_attention.dense.bias": "pytorch_model-17-of-20.bin", "transformer.h.58.post_attention_layernorm.weight": "pytorch_model-17-of-20.bin", "transformer.h.58.post_attention_layernorm.bias": "pytorch_model-17-of-20.bin", "transformer.h.58.mlp.dense_h_to_4h.weight": "pytorch_model-17-of-20.bin", "transformer.h.58.mlp.dense_h_to_4h.bias": "pytorch_model-17-of-20.bin", "transformer.h.58.mlp.dense_4h_to_h.weight": "pytorch_model-17-of-20.bin", "transformer.h.58.mlp.dense_4h_to_h.bias": "pytorch_model-17-of-20.bin", "transformer.h.59.input_layernorm.weight": "pytorch_model-17-of-20.bin", "transformer.h.59.input_layernorm.bias": "pytorch_model-17-of-20.bin", "transformer.h.59.self_attention.query_key_value.weight": "pytorch_model-17-of-20.bin", "transformer.h.59.self_attention.query_key_value.bias": "pytorch_model-17-of-20.bin", "transformer.h.59.self_attention.dense.weight": "pytorch_model-17-of-20.bin", "transformer.h.59.self_attention.dense.bias": "pytorch_model-17-of-20.bin", "transformer.h.59.post_attention_layernorm.weight": "pytorch_model-17-of-20.bin", "transformer.h.59.post_attention_layernorm.bias": "pytorch_model-17-of-20.bin", "transformer.h.59.mlp.dense_h_to_4h.weight": "pytorch_model-17-of-20.bin", "transformer.h.59.mlp.dense_h_to_4h.bias": "pytorch_model-17-of-20.bin", "transformer.h.59.mlp.dense_4h_to_h.weight": "pytorch_model-17-of-20.bin", "transformer.h.59.mlp.dense_4h_to_h.bias": "pytorch_model-17-of-20.bin", "transformer.h.60.input_layernorm.weight": "pytorch_model-17-of-20.bin", "transformer.h.60.input_layernorm.bias": "pytorch_model-17-of-20.bin", "transformer.h.60.self_attention.query_key_value.weight": "pytorch_model-17-of-20.bin", "transformer.h.60.self_attention.query_key_value.bias": "pytorch_model-17-of-20.bin", "transformer.h.60.self_attention.dense.weight": "pytorch_model-17-of-20.bin", "transformer.h.60.self_attention.dense.bias": "pytorch_model-17-of-20.bin", "transformer.h.60.post_attention_layernorm.weight": "pytorch_model-17-of-20.bin", "transformer.h.60.post_attention_layernorm.bias": "pytorch_model-17-of-20.bin", "transformer.h.60.mlp.dense_h_to_4h.weight": "pytorch_model-17-of-20.bin", "transformer.h.60.mlp.dense_h_to_4h.bias": "pytorch_model-17-of-20.bin", "transformer.h.60.mlp.dense_4h_to_h.weight": "pytorch_model-18-of-20.bin", "transformer.h.60.mlp.dense_4h_to_h.bias": "pytorch_model-18-of-20.bin", "transformer.h.61.input_layernorm.weight": "pytorch_model-18-of-20.bin", "transformer.h.61.input_layernorm.bias": "pytorch_model-18-of-20.bin", "transformer.h.61.self_attention.query_key_value.weight": "pytorch_model-18-of-20.bin", "transformer.h.61.self_attention.query_key_value.bias": "pytorch_model-18-of-20.bin", "transformer.h.61.self_attention.dense.weight": "pytorch_model-18-of-20.bin", "transformer.h.61.self_attention.dense.bias": "pytorch_model-18-of-20.bin", "transformer.h.61.post_attention_layernorm.weight": "pytorch_model-18-of-20.bin", "transformer.h.61.post_attention_layernorm.bias": "pytorch_model-18-of-20.bin", "transformer.h.61.mlp.dense_h_to_4h.weight": "pytorch_model-18-of-20.bin", "transformer.h.61.mlp.dense_h_to_4h.bias": "pytorch_model-18-of-20.bin", "transformer.h.61.mlp.dense_4h_to_h.weight": "pytorch_model-18-of-20.bin", "transformer.h.61.mlp.dense_4h_to_h.bias": "pytorch_model-18-of-20.bin", "transformer.h.62.input_layernorm.weight": "pytorch_model-18-of-20.bin", "transformer.h.62.input_layernorm.bias": "pytorch_model-18-of-20.bin", "transformer.h.62.self_attention.query_key_value.weight": "pytorch_model-18-of-20.bin", "transformer.h.62.self_attention.query_key_value.bias": "pytorch_model-18-of-20.bin", "transformer.h.62.self_attention.dense.weight": "pytorch_model-18-of-20.bin", "transformer.h.62.self_attention.dense.bias": "pytorch_model-18-of-20.bin", "transformer.h.62.post_attention_layernorm.weight": "pytorch_model-18-of-20.bin", "transformer.h.62.post_attention_layernorm.bias": "pytorch_model-18-of-20.bin", "transformer.h.62.mlp.dense_h_to_4h.weight": "pytorch_model-18-of-20.bin", "transformer.h.62.mlp.dense_h_to_4h.bias": "pytorch_model-18-of-20.bin", "transformer.h.62.mlp.dense_4h_to_h.weight": "pytorch_model-18-of-20.bin", "transformer.h.62.mlp.dense_4h_to_h.bias": "pytorch_model-18-of-20.bin", "transformer.h.63.input_layernorm.weight": "pytorch_model-18-of-20.bin", "transformer.h.63.input_layernorm.bias": "pytorch_model-18-of-20.bin", "transformer.h.63.self_attention.query_key_value.weight": "pytorch_model-18-of-20.bin", "transformer.h.63.self_attention.query_key_value.bias": "pytorch_model-18-of-20.bin", "transformer.h.63.self_attention.dense.weight": "pytorch_model-18-of-20.bin", "transformer.h.63.self_attention.dense.bias": "pytorch_model-18-of-20.bin", "transformer.h.63.post_attention_layernorm.weight": "pytorch_model-18-of-20.bin", "transformer.h.63.post_attention_layernorm.bias": "pytorch_model-18-of-20.bin", "transformer.h.63.mlp.dense_h_to_4h.weight": "pytorch_model-18-of-20.bin", "transformer.h.63.mlp.dense_h_to_4h.bias": "pytorch_model-18-of-20.bin", "transformer.h.63.mlp.dense_4h_to_h.weight": "pytorch_model-18-of-20.bin", "transformer.h.63.mlp.dense_4h_to_h.bias": "pytorch_model-18-of-20.bin", "transformer.h.64.input_layernorm.weight": "pytorch_model-18-of-20.bin", "transformer.h.64.input_layernorm.bias": "pytorch_model-18-of-20.bin", "transformer.h.64.self_attention.query_key_value.weight": "pytorch_model-18-of-20.bin", "transformer.h.64.self_attention.query_key_value.bias": "pytorch_model-18-of-20.bin", "transformer.h.64.self_attention.dense.weight": "pytorch_model-18-of-20.bin", "transformer.h.64.self_attention.dense.bias": "pytorch_model-18-of-20.bin", "transformer.h.64.post_attention_layernorm.weight": "pytorch_model-18-of-20.bin", "transformer.h.64.post_attention_layernorm.bias": "pytorch_model-18-of-20.bin", "transformer.h.64.mlp.dense_h_to_4h.weight": "pytorch_model-19-of-20.bin", "transformer.h.64.mlp.dense_h_to_4h.bias": "pytorch_model-19-of-20.bin", "transformer.h.64.mlp.dense_4h_to_h.weight": "pytorch_model-19-of-20.bin", "transformer.h.64.mlp.dense_4h_to_h.bias": "pytorch_model-19-of-20.bin", "transformer.h.65.input_layernorm.weight": "pytorch_model-19-of-20.bin", "transformer.h.65.input_layernorm.bias": "pytorch_model-19-of-20.bin", "transformer.h.65.self_attention.query_key_value.weight": "pytorch_model-19-of-20.bin", "transformer.h.65.self_attention.query_key_value.bias": "pytorch_model-19-of-20.bin", "transformer.h.65.self_attention.dense.weight": "pytorch_model-19-of-20.bin", "transformer.h.65.self_attention.dense.bias": "pytorch_model-19-of-20.bin", "transformer.h.65.post_attention_layernorm.weight": "pytorch_model-19-of-20.bin", "transformer.h.65.post_attention_layernorm.bias": "pytorch_model-19-of-20.bin", "transformer.h.65.mlp.dense_h_to_4h.weight": "pytorch_model-19-of-20.bin", "transformer.h.65.mlp.dense_h_to_4h.bias": "pytorch_model-19-of-20.bin", "transformer.h.65.mlp.dense_4h_to_h.weight": "pytorch_model-19-of-20.bin", "transformer.h.65.mlp.dense_4h_to_h.bias": "pytorch_model-19-of-20.bin", "transformer.h.66.input_layernorm.weight": "pytorch_model-19-of-20.bin", "transformer.h.66.input_layernorm.bias": "pytorch_model-19-of-20.bin", "transformer.h.66.self_attention.query_key_value.weight": "pytorch_model-19-of-20.bin", "transformer.h.66.self_attention.query_key_value.bias": "pytorch_model-19-of-20.bin", "transformer.h.66.self_attention.dense.weight": "pytorch_model-19-of-20.bin", "transformer.h.66.self_attention.dense.bias": "pytorch_model-19-of-20.bin", "transformer.h.66.post_attention_layernorm.weight": "pytorch_model-19-of-20.bin", "transformer.h.66.post_attention_layernorm.bias": "pytorch_model-19-of-20.bin", "transformer.h.66.mlp.dense_h_to_4h.weight": "pytorch_model-19-of-20.bin", "transformer.h.66.mlp.dense_h_to_4h.bias": "pytorch_model-19-of-20.bin", "transformer.h.66.mlp.dense_4h_to_h.weight": "pytorch_model-19-of-20.bin", "transformer.h.66.mlp.dense_4h_to_h.bias": "pytorch_model-19-of-20.bin", "transformer.h.67.input_layernorm.weight": "pytorch_model-19-of-20.bin", "transformer.h.67.input_layernorm.bias": "pytorch_model-19-of-20.bin", "transformer.h.67.self_attention.query_key_value.weight": "pytorch_model-19-of-20.bin", "transformer.h.67.self_attention.query_key_value.bias": "pytorch_model-19-of-20.bin", "transformer.h.67.self_attention.dense.weight": "pytorch_model-19-of-20.bin", "transformer.h.67.self_attention.dense.bias": "pytorch_model-19-of-20.bin", "transformer.h.67.post_attention_layernorm.weight": "pytorch_model-19-of-20.bin", "transformer.h.67.post_attention_layernorm.bias": "pytorch_model-19-of-20.bin", "transformer.h.67.mlp.dense_h_to_4h.weight": "pytorch_model-19-of-20.bin", "transformer.h.67.mlp.dense_h_to_4h.bias": "pytorch_model-19-of-20.bin", "transformer.h.67.mlp.dense_4h_to_h.weight": "pytorch_model-19-of-20.bin", "transformer.h.67.mlp.dense_4h_to_h.bias": "pytorch_model-19-of-20.bin", "transformer.h.68.input_layernorm.weight": "pytorch_model-19-of-20.bin", "transformer.h.68.input_layernorm.bias": "pytorch_model-19-of-20.bin", "transformer.h.68.self_attention.query_key_value.weight": "pytorch_model-20-of-20.bin", "transformer.h.68.self_attention.query_key_value.bias": "pytorch_model-20-of-20.bin", "transformer.h.68.self_attention.dense.weight": "pytorch_model-20-of-20.bin", "transformer.h.68.self_attention.dense.bias": "pytorch_model-20-of-20.bin", "transformer.h.68.post_attention_layernorm.weight": "pytorch_model-20-of-20.bin", "transformer.h.68.post_attention_layernorm.bias": "pytorch_model-20-of-20.bin", "transformer.h.68.mlp.dense_h_to_4h.weight": "pytorch_model-20-of-20.bin", "transformer.h.68.mlp.dense_h_to_4h.bias": "pytorch_model-20-of-20.bin", "transformer.h.68.mlp.dense_4h_to_h.weight": "pytorch_model-20-of-20.bin", "transformer.h.68.mlp.dense_4h_to_h.bias": "pytorch_model-20-of-20.bin", "transformer.h.69.input_layernorm.weight": "pytorch_model-20-of-20.bin", "transformer.h.69.input_layernorm.bias": "pytorch_model-20-of-20.bin", "transformer.h.69.self_attention.query_key_value.weight": "pytorch_model-20-of-20.bin", "transformer.h.69.self_attention.query_key_value.bias": "pytorch_model-20-of-20.bin", "transformer.h.69.self_attention.dense.weight": "pytorch_model-20-of-20.bin", "transformer.h.69.self_attention.dense.bias": "pytorch_model-20-of-20.bin", "transformer.h.69.post_attention_layernorm.weight": "pytorch_model-20-of-20.bin", "transformer.h.69.post_attention_layernorm.bias": "pytorch_model-20-of-20.bin", "transformer.h.69.mlp.dense_h_to_4h.weight": "pytorch_model-20-of-20.bin", "transformer.h.69.mlp.dense_h_to_4h.bias": "pytorch_model-20-of-20.bin", "transformer.h.69.mlp.dense_4h_to_h.weight": "pytorch_model-20-of-20.bin", "transformer.h.69.mlp.dense_4h_to_h.bias": "pytorch_model-20-of-20.bin", "transformer.ln_f.weight": "pytorch_model-20-of-20.bin", "transformer.ln_f.bias": "pytorch_model-20-of-20.bin", "lm_head.weight": "pytorch_model-20-of-20.bin"}}
pytorch_model.bin.sambatensor_index.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"transformer.word_embeddings.weight": "bloomforcausallm__transformer__word_embeddings__weight", "transformer.word_embeddings_layernorm.weight": "bloomforcausallm__transformer__word_embeddings_layernorm__weight", "transformer.word_embeddings_layernorm.bias": "bloomforcausallm__transformer__word_embeddings_layernorm__bias", "transformer.h.0.input_layernorm.weight": "bloomforcausallm__transformer__h__0__input_layernorm__weight", "transformer.h.0.input_layernorm.bias": "bloomforcausallm__transformer__h__0__input_layernorm__bias", "transformer.h.0.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__0__self_attention__query_key_value__weight", "transformer.h.0.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__0__self_attention__query_key_value__bias", "transformer.h.0.self_attention.dense.weight": "bloomforcausallm__transformer__h__0__self_attention__dense__weight", "transformer.h.0.self_attention.dense.bias": "bloomforcausallm__transformer__h__0__self_attention__dense__bias", "transformer.h.0.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__0__post_attention_layernorm__weight", "transformer.h.0.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__0__post_attention_layernorm__bias", "transformer.h.0.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__0__mlp__dense_h_to_4h__weight", "transformer.h.0.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__0__mlp__dense_h_to_4h__bias", "transformer.h.0.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__0__mlp__dense_4h_to_h__weight", "transformer.h.0.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__0__mlp__dense_4h_to_h__bias", "transformer.h.1.input_layernorm.weight": "bloomforcausallm__transformer__h__1__input_layernorm__weight", "transformer.h.1.input_layernorm.bias": "bloomforcausallm__transformer__h__1__input_layernorm__bias", "transformer.h.1.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__1__self_attention__query_key_value__weight", "transformer.h.1.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__1__self_attention__query_key_value__bias", "transformer.h.1.self_attention.dense.weight": "bloomforcausallm__transformer__h__1__self_attention__dense__weight", "transformer.h.1.self_attention.dense.bias": "bloomforcausallm__transformer__h__1__self_attention__dense__bias", "transformer.h.1.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__1__post_attention_layernorm__weight", "transformer.h.1.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__1__post_attention_layernorm__bias", "transformer.h.1.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__1__mlp__dense_h_to_4h__weight", "transformer.h.1.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__1__mlp__dense_h_to_4h__bias", "transformer.h.1.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__1__mlp__dense_4h_to_h__weight", "transformer.h.1.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__1__mlp__dense_4h_to_h__bias", "transformer.h.2.input_layernorm.weight": "bloomforcausallm__transformer__h__2__input_layernorm__weight", "transformer.h.2.input_layernorm.bias": "bloomforcausallm__transformer__h__2__input_layernorm__bias", "transformer.h.2.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__2__self_attention__query_key_value__weight", "transformer.h.2.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__2__self_attention__query_key_value__bias", "transformer.h.2.self_attention.dense.weight": "bloomforcausallm__transformer__h__2__self_attention__dense__weight", "transformer.h.2.self_attention.dense.bias": "bloomforcausallm__transformer__h__2__self_attention__dense__bias", "transformer.h.2.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__2__post_attention_layernorm__weight", "transformer.h.2.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__2__post_attention_layernorm__bias", "transformer.h.2.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__2__mlp__dense_h_to_4h__weight", "transformer.h.2.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__2__mlp__dense_h_to_4h__bias", "transformer.h.2.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__2__mlp__dense_4h_to_h__weight", "transformer.h.2.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__2__mlp__dense_4h_to_h__bias", "transformer.h.3.input_layernorm.weight": "bloomforcausallm__transformer__h__3__input_layernorm__weight", "transformer.h.3.input_layernorm.bias": "bloomforcausallm__transformer__h__3__input_layernorm__bias", "transformer.h.3.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__3__self_attention__query_key_value__weight", "transformer.h.3.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__3__self_attention__query_key_value__bias", "transformer.h.3.self_attention.dense.weight": "bloomforcausallm__transformer__h__3__self_attention__dense__weight", "transformer.h.3.self_attention.dense.bias": "bloomforcausallm__transformer__h__3__self_attention__dense__bias", "transformer.h.3.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__3__post_attention_layernorm__weight", "transformer.h.3.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__3__post_attention_layernorm__bias", "transformer.h.3.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__3__mlp__dense_h_to_4h__weight", "transformer.h.3.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__3__mlp__dense_h_to_4h__bias", "transformer.h.3.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__3__mlp__dense_4h_to_h__weight", "transformer.h.3.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__3__mlp__dense_4h_to_h__bias", "transformer.h.4.input_layernorm.weight": "bloomforcausallm__transformer__h__4__input_layernorm__weight", "transformer.h.4.input_layernorm.bias": "bloomforcausallm__transformer__h__4__input_layernorm__bias", "transformer.h.4.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__4__self_attention__query_key_value__weight", "transformer.h.4.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__4__self_attention__query_key_value__bias", "transformer.h.4.self_attention.dense.weight": "bloomforcausallm__transformer__h__4__self_attention__dense__weight", "transformer.h.4.self_attention.dense.bias": "bloomforcausallm__transformer__h__4__self_attention__dense__bias", "transformer.h.4.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__4__post_attention_layernorm__weight", "transformer.h.4.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__4__post_attention_layernorm__bias", "transformer.h.4.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__4__mlp__dense_h_to_4h__weight", "transformer.h.4.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__4__mlp__dense_h_to_4h__bias", "transformer.h.4.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__4__mlp__dense_4h_to_h__weight", "transformer.h.4.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__4__mlp__dense_4h_to_h__bias", "transformer.h.5.input_layernorm.weight": "bloomforcausallm__transformer__h__5__input_layernorm__weight", "transformer.h.5.input_layernorm.bias": "bloomforcausallm__transformer__h__5__input_layernorm__bias", "transformer.h.5.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__5__self_attention__query_key_value__weight", "transformer.h.5.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__5__self_attention__query_key_value__bias", "transformer.h.5.self_attention.dense.weight": "bloomforcausallm__transformer__h__5__self_attention__dense__weight", "transformer.h.5.self_attention.dense.bias": "bloomforcausallm__transformer__h__5__self_attention__dense__bias", "transformer.h.5.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__5__post_attention_layernorm__weight", "transformer.h.5.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__5__post_attention_layernorm__bias", "transformer.h.5.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__5__mlp__dense_h_to_4h__weight", "transformer.h.5.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__5__mlp__dense_h_to_4h__bias", "transformer.h.5.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__5__mlp__dense_4h_to_h__weight", "transformer.h.5.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__5__mlp__dense_4h_to_h__bias", "transformer.h.6.input_layernorm.weight": "bloomforcausallm__transformer__h__6__input_layernorm__weight", "transformer.h.6.input_layernorm.bias": "bloomforcausallm__transformer__h__6__input_layernorm__bias", "transformer.h.6.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__6__self_attention__query_key_value__weight", "transformer.h.6.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__6__self_attention__query_key_value__bias", "transformer.h.6.self_attention.dense.weight": "bloomforcausallm__transformer__h__6__self_attention__dense__weight", "transformer.h.6.self_attention.dense.bias": "bloomforcausallm__transformer__h__6__self_attention__dense__bias", "transformer.h.6.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__6__post_attention_layernorm__weight", "transformer.h.6.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__6__post_attention_layernorm__bias", "transformer.h.6.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__6__mlp__dense_h_to_4h__weight", "transformer.h.6.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__6__mlp__dense_h_to_4h__bias", "transformer.h.6.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__6__mlp__dense_4h_to_h__weight", "transformer.h.6.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__6__mlp__dense_4h_to_h__bias", "transformer.h.7.input_layernorm.weight": "bloomforcausallm__transformer__h__7__input_layernorm__weight", "transformer.h.7.input_layernorm.bias": "bloomforcausallm__transformer__h__7__input_layernorm__bias", "transformer.h.7.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__7__self_attention__query_key_value__weight", "transformer.h.7.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__7__self_attention__query_key_value__bias", "transformer.h.7.self_attention.dense.weight": "bloomforcausallm__transformer__h__7__self_attention__dense__weight", "transformer.h.7.self_attention.dense.bias": "bloomforcausallm__transformer__h__7__self_attention__dense__bias", "transformer.h.7.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__7__post_attention_layernorm__weight", "transformer.h.7.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__7__post_attention_layernorm__bias", "transformer.h.7.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__7__mlp__dense_h_to_4h__weight", "transformer.h.7.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__7__mlp__dense_h_to_4h__bias", "transformer.h.7.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__7__mlp__dense_4h_to_h__weight", "transformer.h.7.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__7__mlp__dense_4h_to_h__bias", "transformer.h.8.input_layernorm.weight": "bloomforcausallm__transformer__h__8__input_layernorm__weight", "transformer.h.8.input_layernorm.bias": "bloomforcausallm__transformer__h__8__input_layernorm__bias", "transformer.h.8.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__8__self_attention__query_key_value__weight", "transformer.h.8.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__8__self_attention__query_key_value__bias", "transformer.h.8.self_attention.dense.weight": "bloomforcausallm__transformer__h__8__self_attention__dense__weight", "transformer.h.8.self_attention.dense.bias": "bloomforcausallm__transformer__h__8__self_attention__dense__bias", "transformer.h.8.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__8__post_attention_layernorm__weight", "transformer.h.8.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__8__post_attention_layernorm__bias", "transformer.h.8.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__8__mlp__dense_h_to_4h__weight", "transformer.h.8.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__8__mlp__dense_h_to_4h__bias", "transformer.h.8.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__8__mlp__dense_4h_to_h__weight", "transformer.h.8.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__8__mlp__dense_4h_to_h__bias", "transformer.h.9.input_layernorm.weight": "bloomforcausallm__transformer__h__9__input_layernorm__weight", "transformer.h.9.input_layernorm.bias": "bloomforcausallm__transformer__h__9__input_layernorm__bias", "transformer.h.9.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__9__self_attention__query_key_value__weight", "transformer.h.9.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__9__self_attention__query_key_value__bias", "transformer.h.9.self_attention.dense.weight": "bloomforcausallm__transformer__h__9__self_attention__dense__weight", "transformer.h.9.self_attention.dense.bias": "bloomforcausallm__transformer__h__9__self_attention__dense__bias", "transformer.h.9.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__9__post_attention_layernorm__weight", "transformer.h.9.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__9__post_attention_layernorm__bias", "transformer.h.9.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__9__mlp__dense_h_to_4h__weight", "transformer.h.9.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__9__mlp__dense_h_to_4h__bias", "transformer.h.9.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__9__mlp__dense_4h_to_h__weight", "transformer.h.9.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__9__mlp__dense_4h_to_h__bias", "transformer.h.10.input_layernorm.weight": "bloomforcausallm__transformer__h__10__input_layernorm__weight", "transformer.h.10.input_layernorm.bias": "bloomforcausallm__transformer__h__10__input_layernorm__bias", "transformer.h.10.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__10__self_attention__query_key_value__weight", "transformer.h.10.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__10__self_attention__query_key_value__bias", "transformer.h.10.self_attention.dense.weight": "bloomforcausallm__transformer__h__10__self_attention__dense__weight", "transformer.h.10.self_attention.dense.bias": "bloomforcausallm__transformer__h__10__self_attention__dense__bias", "transformer.h.10.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__10__post_attention_layernorm__weight", "transformer.h.10.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__10__post_attention_layernorm__bias", "transformer.h.10.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__10__mlp__dense_h_to_4h__weight", "transformer.h.10.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__10__mlp__dense_h_to_4h__bias", "transformer.h.10.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__10__mlp__dense_4h_to_h__weight", "transformer.h.10.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__10__mlp__dense_4h_to_h__bias", "transformer.h.11.input_layernorm.weight": "bloomforcausallm__transformer__h__11__input_layernorm__weight", "transformer.h.11.input_layernorm.bias": "bloomforcausallm__transformer__h__11__input_layernorm__bias", "transformer.h.11.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__11__self_attention__query_key_value__weight", "transformer.h.11.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__11__self_attention__query_key_value__bias", "transformer.h.11.self_attention.dense.weight": "bloomforcausallm__transformer__h__11__self_attention__dense__weight", "transformer.h.11.self_attention.dense.bias": "bloomforcausallm__transformer__h__11__self_attention__dense__bias", "transformer.h.11.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__11__post_attention_layernorm__weight", "transformer.h.11.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__11__post_attention_layernorm__bias", "transformer.h.11.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__11__mlp__dense_h_to_4h__weight", "transformer.h.11.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__11__mlp__dense_h_to_4h__bias", "transformer.h.11.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__11__mlp__dense_4h_to_h__weight", "transformer.h.11.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__11__mlp__dense_4h_to_h__bias", "transformer.h.12.input_layernorm.weight": "bloomforcausallm__transformer__h__12__input_layernorm__weight", "transformer.h.12.input_layernorm.bias": "bloomforcausallm__transformer__h__12__input_layernorm__bias", "transformer.h.12.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__12__self_attention__query_key_value__weight", "transformer.h.12.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__12__self_attention__query_key_value__bias", "transformer.h.12.self_attention.dense.weight": "bloomforcausallm__transformer__h__12__self_attention__dense__weight", "transformer.h.12.self_attention.dense.bias": "bloomforcausallm__transformer__h__12__self_attention__dense__bias", "transformer.h.12.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__12__post_attention_layernorm__weight", "transformer.h.12.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__12__post_attention_layernorm__bias", "transformer.h.12.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__12__mlp__dense_h_to_4h__weight", "transformer.h.12.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__12__mlp__dense_h_to_4h__bias", "transformer.h.12.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__12__mlp__dense_4h_to_h__weight", "transformer.h.12.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__12__mlp__dense_4h_to_h__bias", "transformer.h.13.input_layernorm.weight": "bloomforcausallm__transformer__h__13__input_layernorm__weight", "transformer.h.13.input_layernorm.bias": "bloomforcausallm__transformer__h__13__input_layernorm__bias", "transformer.h.13.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__13__self_attention__query_key_value__weight", "transformer.h.13.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__13__self_attention__query_key_value__bias", "transformer.h.13.self_attention.dense.weight": "bloomforcausallm__transformer__h__13__self_attention__dense__weight", "transformer.h.13.self_attention.dense.bias": "bloomforcausallm__transformer__h__13__self_attention__dense__bias", "transformer.h.13.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__13__post_attention_layernorm__weight", "transformer.h.13.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__13__post_attention_layernorm__bias", "transformer.h.13.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__13__mlp__dense_h_to_4h__weight", "transformer.h.13.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__13__mlp__dense_h_to_4h__bias", "transformer.h.13.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__13__mlp__dense_4h_to_h__weight", "transformer.h.13.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__13__mlp__dense_4h_to_h__bias", "transformer.h.14.input_layernorm.weight": "bloomforcausallm__transformer__h__14__input_layernorm__weight", "transformer.h.14.input_layernorm.bias": "bloomforcausallm__transformer__h__14__input_layernorm__bias", "transformer.h.14.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__14__self_attention__query_key_value__weight", "transformer.h.14.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__14__self_attention__query_key_value__bias", "transformer.h.14.self_attention.dense.weight": "bloomforcausallm__transformer__h__14__self_attention__dense__weight", "transformer.h.14.self_attention.dense.bias": "bloomforcausallm__transformer__h__14__self_attention__dense__bias", "transformer.h.14.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__14__post_attention_layernorm__weight", "transformer.h.14.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__14__post_attention_layernorm__bias", "transformer.h.14.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__14__mlp__dense_h_to_4h__weight", "transformer.h.14.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__14__mlp__dense_h_to_4h__bias", "transformer.h.14.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__14__mlp__dense_4h_to_h__weight", "transformer.h.14.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__14__mlp__dense_4h_to_h__bias", "transformer.h.15.input_layernorm.weight": "bloomforcausallm__transformer__h__15__input_layernorm__weight", "transformer.h.15.input_layernorm.bias": "bloomforcausallm__transformer__h__15__input_layernorm__bias", "transformer.h.15.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__15__self_attention__query_key_value__weight", "transformer.h.15.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__15__self_attention__query_key_value__bias", "transformer.h.15.self_attention.dense.weight": "bloomforcausallm__transformer__h__15__self_attention__dense__weight", "transformer.h.15.self_attention.dense.bias": "bloomforcausallm__transformer__h__15__self_attention__dense__bias", "transformer.h.15.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__15__post_attention_layernorm__weight", "transformer.h.15.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__15__post_attention_layernorm__bias", "transformer.h.15.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__15__mlp__dense_h_to_4h__weight", "transformer.h.15.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__15__mlp__dense_h_to_4h__bias", "transformer.h.15.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__15__mlp__dense_4h_to_h__weight", "transformer.h.15.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__15__mlp__dense_4h_to_h__bias", "transformer.h.16.input_layernorm.weight": "bloomforcausallm__transformer__h__16__input_layernorm__weight", "transformer.h.16.input_layernorm.bias": "bloomforcausallm__transformer__h__16__input_layernorm__bias", "transformer.h.16.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__16__self_attention__query_key_value__weight", "transformer.h.16.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__16__self_attention__query_key_value__bias", "transformer.h.16.self_attention.dense.weight": "bloomforcausallm__transformer__h__16__self_attention__dense__weight", "transformer.h.16.self_attention.dense.bias": "bloomforcausallm__transformer__h__16__self_attention__dense__bias", "transformer.h.16.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__16__post_attention_layernorm__weight", "transformer.h.16.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__16__post_attention_layernorm__bias", "transformer.h.16.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__16__mlp__dense_h_to_4h__weight", "transformer.h.16.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__16__mlp__dense_h_to_4h__bias", "transformer.h.16.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__16__mlp__dense_4h_to_h__weight", "transformer.h.16.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__16__mlp__dense_4h_to_h__bias", "transformer.h.17.input_layernorm.weight": "bloomforcausallm__transformer__h__17__input_layernorm__weight", "transformer.h.17.input_layernorm.bias": "bloomforcausallm__transformer__h__17__input_layernorm__bias", "transformer.h.17.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__17__self_attention__query_key_value__weight", "transformer.h.17.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__17__self_attention__query_key_value__bias", "transformer.h.17.self_attention.dense.weight": "bloomforcausallm__transformer__h__17__self_attention__dense__weight", "transformer.h.17.self_attention.dense.bias": "bloomforcausallm__transformer__h__17__self_attention__dense__bias", "transformer.h.17.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__17__post_attention_layernorm__weight", "transformer.h.17.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__17__post_attention_layernorm__bias", "transformer.h.17.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__17__mlp__dense_h_to_4h__weight", "transformer.h.17.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__17__mlp__dense_h_to_4h__bias", "transformer.h.17.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__17__mlp__dense_4h_to_h__weight", "transformer.h.17.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__17__mlp__dense_4h_to_h__bias", "transformer.h.18.input_layernorm.weight": "bloomforcausallm__transformer__h__18__input_layernorm__weight", "transformer.h.18.input_layernorm.bias": "bloomforcausallm__transformer__h__18__input_layernorm__bias", "transformer.h.18.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__18__self_attention__query_key_value__weight", "transformer.h.18.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__18__self_attention__query_key_value__bias", "transformer.h.18.self_attention.dense.weight": "bloomforcausallm__transformer__h__18__self_attention__dense__weight", "transformer.h.18.self_attention.dense.bias": "bloomforcausallm__transformer__h__18__self_attention__dense__bias", "transformer.h.18.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__18__post_attention_layernorm__weight", "transformer.h.18.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__18__post_attention_layernorm__bias", "transformer.h.18.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__18__mlp__dense_h_to_4h__weight", "transformer.h.18.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__18__mlp__dense_h_to_4h__bias", "transformer.h.18.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__18__mlp__dense_4h_to_h__weight", "transformer.h.18.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__18__mlp__dense_4h_to_h__bias", "transformer.h.19.input_layernorm.weight": "bloomforcausallm__transformer__h__19__input_layernorm__weight", "transformer.h.19.input_layernorm.bias": "bloomforcausallm__transformer__h__19__input_layernorm__bias", "transformer.h.19.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__19__self_attention__query_key_value__weight", "transformer.h.19.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__19__self_attention__query_key_value__bias", "transformer.h.19.self_attention.dense.weight": "bloomforcausallm__transformer__h__19__self_attention__dense__weight", "transformer.h.19.self_attention.dense.bias": "bloomforcausallm__transformer__h__19__self_attention__dense__bias", "transformer.h.19.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__19__post_attention_layernorm__weight", "transformer.h.19.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__19__post_attention_layernorm__bias", "transformer.h.19.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__19__mlp__dense_h_to_4h__weight", "transformer.h.19.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__19__mlp__dense_h_to_4h__bias", "transformer.h.19.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__19__mlp__dense_4h_to_h__weight", "transformer.h.19.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__19__mlp__dense_4h_to_h__bias", "transformer.h.20.input_layernorm.weight": "bloomforcausallm__transformer__h__20__input_layernorm__weight", "transformer.h.20.input_layernorm.bias": "bloomforcausallm__transformer__h__20__input_layernorm__bias", "transformer.h.20.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__20__self_attention__query_key_value__weight", "transformer.h.20.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__20__self_attention__query_key_value__bias", "transformer.h.20.self_attention.dense.weight": "bloomforcausallm__transformer__h__20__self_attention__dense__weight", "transformer.h.20.self_attention.dense.bias": "bloomforcausallm__transformer__h__20__self_attention__dense__bias", "transformer.h.20.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__20__post_attention_layernorm__weight", "transformer.h.20.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__20__post_attention_layernorm__bias", "transformer.h.20.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__20__mlp__dense_h_to_4h__weight", "transformer.h.20.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__20__mlp__dense_h_to_4h__bias", "transformer.h.20.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__20__mlp__dense_4h_to_h__weight", "transformer.h.20.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__20__mlp__dense_4h_to_h__bias", "transformer.h.21.input_layernorm.weight": "bloomforcausallm__transformer__h__21__input_layernorm__weight", "transformer.h.21.input_layernorm.bias": "bloomforcausallm__transformer__h__21__input_layernorm__bias", "transformer.h.21.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__21__self_attention__query_key_value__weight", "transformer.h.21.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__21__self_attention__query_key_value__bias", "transformer.h.21.self_attention.dense.weight": "bloomforcausallm__transformer__h__21__self_attention__dense__weight", "transformer.h.21.self_attention.dense.bias": "bloomforcausallm__transformer__h__21__self_attention__dense__bias", "transformer.h.21.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__21__post_attention_layernorm__weight", "transformer.h.21.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__21__post_attention_layernorm__bias", "transformer.h.21.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__21__mlp__dense_h_to_4h__weight", "transformer.h.21.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__21__mlp__dense_h_to_4h__bias", "transformer.h.21.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__21__mlp__dense_4h_to_h__weight", "transformer.h.21.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__21__mlp__dense_4h_to_h__bias", "transformer.h.22.input_layernorm.weight": "bloomforcausallm__transformer__h__22__input_layernorm__weight", "transformer.h.22.input_layernorm.bias": "bloomforcausallm__transformer__h__22__input_layernorm__bias", "transformer.h.22.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__22__self_attention__query_key_value__weight", "transformer.h.22.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__22__self_attention__query_key_value__bias", "transformer.h.22.self_attention.dense.weight": "bloomforcausallm__transformer__h__22__self_attention__dense__weight", "transformer.h.22.self_attention.dense.bias": "bloomforcausallm__transformer__h__22__self_attention__dense__bias", "transformer.h.22.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__22__post_attention_layernorm__weight", "transformer.h.22.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__22__post_attention_layernorm__bias", "transformer.h.22.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__22__mlp__dense_h_to_4h__weight", "transformer.h.22.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__22__mlp__dense_h_to_4h__bias", "transformer.h.22.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__22__mlp__dense_4h_to_h__weight", "transformer.h.22.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__22__mlp__dense_4h_to_h__bias", "transformer.h.23.input_layernorm.weight": "bloomforcausallm__transformer__h__23__input_layernorm__weight", "transformer.h.23.input_layernorm.bias": "bloomforcausallm__transformer__h__23__input_layernorm__bias", "transformer.h.23.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__23__self_attention__query_key_value__weight", "transformer.h.23.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__23__self_attention__query_key_value__bias", "transformer.h.23.self_attention.dense.weight": "bloomforcausallm__transformer__h__23__self_attention__dense__weight", "transformer.h.23.self_attention.dense.bias": "bloomforcausallm__transformer__h__23__self_attention__dense__bias", "transformer.h.23.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__23__post_attention_layernorm__weight", "transformer.h.23.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__23__post_attention_layernorm__bias", "transformer.h.23.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__23__mlp__dense_h_to_4h__weight", "transformer.h.23.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__23__mlp__dense_h_to_4h__bias", "transformer.h.23.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__23__mlp__dense_4h_to_h__weight", "transformer.h.23.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__23__mlp__dense_4h_to_h__bias", "transformer.h.24.input_layernorm.weight": "bloomforcausallm__transformer__h__24__input_layernorm__weight", "transformer.h.24.input_layernorm.bias": "bloomforcausallm__transformer__h__24__input_layernorm__bias", "transformer.h.24.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__24__self_attention__query_key_value__weight", "transformer.h.24.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__24__self_attention__query_key_value__bias", "transformer.h.24.self_attention.dense.weight": "bloomforcausallm__transformer__h__24__self_attention__dense__weight", "transformer.h.24.self_attention.dense.bias": "bloomforcausallm__transformer__h__24__self_attention__dense__bias", "transformer.h.24.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__24__post_attention_layernorm__weight", "transformer.h.24.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__24__post_attention_layernorm__bias", "transformer.h.24.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__24__mlp__dense_h_to_4h__weight", "transformer.h.24.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__24__mlp__dense_h_to_4h__bias", "transformer.h.24.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__24__mlp__dense_4h_to_h__weight", "transformer.h.24.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__24__mlp__dense_4h_to_h__bias", "transformer.h.25.input_layernorm.weight": "bloomforcausallm__transformer__h__25__input_layernorm__weight", "transformer.h.25.input_layernorm.bias": "bloomforcausallm__transformer__h__25__input_layernorm__bias", "transformer.h.25.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__25__self_attention__query_key_value__weight", "transformer.h.25.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__25__self_attention__query_key_value__bias", "transformer.h.25.self_attention.dense.weight": "bloomforcausallm__transformer__h__25__self_attention__dense__weight", "transformer.h.25.self_attention.dense.bias": "bloomforcausallm__transformer__h__25__self_attention__dense__bias", "transformer.h.25.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__25__post_attention_layernorm__weight", "transformer.h.25.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__25__post_attention_layernorm__bias", "transformer.h.25.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__25__mlp__dense_h_to_4h__weight", "transformer.h.25.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__25__mlp__dense_h_to_4h__bias", "transformer.h.25.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__25__mlp__dense_4h_to_h__weight", "transformer.h.25.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__25__mlp__dense_4h_to_h__bias", "transformer.h.26.input_layernorm.weight": "bloomforcausallm__transformer__h__26__input_layernorm__weight", "transformer.h.26.input_layernorm.bias": "bloomforcausallm__transformer__h__26__input_layernorm__bias", "transformer.h.26.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__26__self_attention__query_key_value__weight", "transformer.h.26.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__26__self_attention__query_key_value__bias", "transformer.h.26.self_attention.dense.weight": "bloomforcausallm__transformer__h__26__self_attention__dense__weight", "transformer.h.26.self_attention.dense.bias": "bloomforcausallm__transformer__h__26__self_attention__dense__bias", "transformer.h.26.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__26__post_attention_layernorm__weight", "transformer.h.26.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__26__post_attention_layernorm__bias", "transformer.h.26.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__26__mlp__dense_h_to_4h__weight", "transformer.h.26.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__26__mlp__dense_h_to_4h__bias", "transformer.h.26.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__26__mlp__dense_4h_to_h__weight", "transformer.h.26.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__26__mlp__dense_4h_to_h__bias", "transformer.h.27.input_layernorm.weight": "bloomforcausallm__transformer__h__27__input_layernorm__weight", "transformer.h.27.input_layernorm.bias": "bloomforcausallm__transformer__h__27__input_layernorm__bias", "transformer.h.27.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__27__self_attention__query_key_value__weight", "transformer.h.27.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__27__self_attention__query_key_value__bias", "transformer.h.27.self_attention.dense.weight": "bloomforcausallm__transformer__h__27__self_attention__dense__weight", "transformer.h.27.self_attention.dense.bias": "bloomforcausallm__transformer__h__27__self_attention__dense__bias", "transformer.h.27.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__27__post_attention_layernorm__weight", "transformer.h.27.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__27__post_attention_layernorm__bias", "transformer.h.27.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__27__mlp__dense_h_to_4h__weight", "transformer.h.27.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__27__mlp__dense_h_to_4h__bias", "transformer.h.27.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__27__mlp__dense_4h_to_h__weight", "transformer.h.27.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__27__mlp__dense_4h_to_h__bias", "transformer.h.28.input_layernorm.weight": "bloomforcausallm__transformer__h__28__input_layernorm__weight", "transformer.h.28.input_layernorm.bias": "bloomforcausallm__transformer__h__28__input_layernorm__bias", "transformer.h.28.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__28__self_attention__query_key_value__weight", "transformer.h.28.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__28__self_attention__query_key_value__bias", "transformer.h.28.self_attention.dense.weight": "bloomforcausallm__transformer__h__28__self_attention__dense__weight", "transformer.h.28.self_attention.dense.bias": "bloomforcausallm__transformer__h__28__self_attention__dense__bias", "transformer.h.28.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__28__post_attention_layernorm__weight", "transformer.h.28.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__28__post_attention_layernorm__bias", "transformer.h.28.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__28__mlp__dense_h_to_4h__weight", "transformer.h.28.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__28__mlp__dense_h_to_4h__bias", "transformer.h.28.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__28__mlp__dense_4h_to_h__weight", "transformer.h.28.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__28__mlp__dense_4h_to_h__bias", "transformer.h.29.input_layernorm.weight": "bloomforcausallm__transformer__h__29__input_layernorm__weight", "transformer.h.29.input_layernorm.bias": "bloomforcausallm__transformer__h__29__input_layernorm__bias", "transformer.h.29.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__29__self_attention__query_key_value__weight", "transformer.h.29.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__29__self_attention__query_key_value__bias", "transformer.h.29.self_attention.dense.weight": "bloomforcausallm__transformer__h__29__self_attention__dense__weight", "transformer.h.29.self_attention.dense.bias": "bloomforcausallm__transformer__h__29__self_attention__dense__bias", "transformer.h.29.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__29__post_attention_layernorm__weight", "transformer.h.29.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__29__post_attention_layernorm__bias", "transformer.h.29.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__29__mlp__dense_h_to_4h__weight", "transformer.h.29.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__29__mlp__dense_h_to_4h__bias", "transformer.h.29.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__29__mlp__dense_4h_to_h__weight", "transformer.h.29.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__29__mlp__dense_4h_to_h__bias", "transformer.h.30.input_layernorm.weight": "bloomforcausallm__transformer__h__30__input_layernorm__weight", "transformer.h.30.input_layernorm.bias": "bloomforcausallm__transformer__h__30__input_layernorm__bias", "transformer.h.30.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__30__self_attention__query_key_value__weight", "transformer.h.30.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__30__self_attention__query_key_value__bias", "transformer.h.30.self_attention.dense.weight": "bloomforcausallm__transformer__h__30__self_attention__dense__weight", "transformer.h.30.self_attention.dense.bias": "bloomforcausallm__transformer__h__30__self_attention__dense__bias", "transformer.h.30.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__30__post_attention_layernorm__weight", "transformer.h.30.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__30__post_attention_layernorm__bias", "transformer.h.30.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__30__mlp__dense_h_to_4h__weight", "transformer.h.30.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__30__mlp__dense_h_to_4h__bias", "transformer.h.30.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__30__mlp__dense_4h_to_h__weight", "transformer.h.30.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__30__mlp__dense_4h_to_h__bias", "transformer.h.31.input_layernorm.weight": "bloomforcausallm__transformer__h__31__input_layernorm__weight", "transformer.h.31.input_layernorm.bias": "bloomforcausallm__transformer__h__31__input_layernorm__bias", "transformer.h.31.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__31__self_attention__query_key_value__weight", "transformer.h.31.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__31__self_attention__query_key_value__bias", "transformer.h.31.self_attention.dense.weight": "bloomforcausallm__transformer__h__31__self_attention__dense__weight", "transformer.h.31.self_attention.dense.bias": "bloomforcausallm__transformer__h__31__self_attention__dense__bias", "transformer.h.31.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__31__post_attention_layernorm__weight", "transformer.h.31.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__31__post_attention_layernorm__bias", "transformer.h.31.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__31__mlp__dense_h_to_4h__weight", "transformer.h.31.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__31__mlp__dense_h_to_4h__bias", "transformer.h.31.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__31__mlp__dense_4h_to_h__weight", "transformer.h.31.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__31__mlp__dense_4h_to_h__bias", "transformer.h.32.input_layernorm.weight": "bloomforcausallm__transformer__h__32__input_layernorm__weight", "transformer.h.32.input_layernorm.bias": "bloomforcausallm__transformer__h__32__input_layernorm__bias", "transformer.h.32.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__32__self_attention__query_key_value__weight", "transformer.h.32.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__32__self_attention__query_key_value__bias", "transformer.h.32.self_attention.dense.weight": "bloomforcausallm__transformer__h__32__self_attention__dense__weight", "transformer.h.32.self_attention.dense.bias": "bloomforcausallm__transformer__h__32__self_attention__dense__bias", "transformer.h.32.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__32__post_attention_layernorm__weight", "transformer.h.32.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__32__post_attention_layernorm__bias", "transformer.h.32.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__32__mlp__dense_h_to_4h__weight", "transformer.h.32.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__32__mlp__dense_h_to_4h__bias", "transformer.h.32.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__32__mlp__dense_4h_to_h__weight", "transformer.h.32.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__32__mlp__dense_4h_to_h__bias", "transformer.h.33.input_layernorm.weight": "bloomforcausallm__transformer__h__33__input_layernorm__weight", "transformer.h.33.input_layernorm.bias": "bloomforcausallm__transformer__h__33__input_layernorm__bias", "transformer.h.33.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__33__self_attention__query_key_value__weight", "transformer.h.33.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__33__self_attention__query_key_value__bias", "transformer.h.33.self_attention.dense.weight": "bloomforcausallm__transformer__h__33__self_attention__dense__weight", "transformer.h.33.self_attention.dense.bias": "bloomforcausallm__transformer__h__33__self_attention__dense__bias", "transformer.h.33.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__33__post_attention_layernorm__weight", "transformer.h.33.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__33__post_attention_layernorm__bias", "transformer.h.33.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__33__mlp__dense_h_to_4h__weight", "transformer.h.33.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__33__mlp__dense_h_to_4h__bias", "transformer.h.33.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__33__mlp__dense_4h_to_h__weight", "transformer.h.33.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__33__mlp__dense_4h_to_h__bias", "transformer.h.34.input_layernorm.weight": "bloomforcausallm__transformer__h__34__input_layernorm__weight", "transformer.h.34.input_layernorm.bias": "bloomforcausallm__transformer__h__34__input_layernorm__bias", "transformer.h.34.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__34__self_attention__query_key_value__weight", "transformer.h.34.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__34__self_attention__query_key_value__bias", "transformer.h.34.self_attention.dense.weight": "bloomforcausallm__transformer__h__34__self_attention__dense__weight", "transformer.h.34.self_attention.dense.bias": "bloomforcausallm__transformer__h__34__self_attention__dense__bias", "transformer.h.34.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__34__post_attention_layernorm__weight", "transformer.h.34.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__34__post_attention_layernorm__bias", "transformer.h.34.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__34__mlp__dense_h_to_4h__weight", "transformer.h.34.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__34__mlp__dense_h_to_4h__bias", "transformer.h.34.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__34__mlp__dense_4h_to_h__weight", "transformer.h.34.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__34__mlp__dense_4h_to_h__bias", "transformer.h.35.input_layernorm.weight": "bloomforcausallm__transformer__h__35__input_layernorm__weight", "transformer.h.35.input_layernorm.bias": "bloomforcausallm__transformer__h__35__input_layernorm__bias", "transformer.h.35.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__35__self_attention__query_key_value__weight", "transformer.h.35.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__35__self_attention__query_key_value__bias", "transformer.h.35.self_attention.dense.weight": "bloomforcausallm__transformer__h__35__self_attention__dense__weight", "transformer.h.35.self_attention.dense.bias": "bloomforcausallm__transformer__h__35__self_attention__dense__bias", "transformer.h.35.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__35__post_attention_layernorm__weight", "transformer.h.35.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__35__post_attention_layernorm__bias", "transformer.h.35.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__35__mlp__dense_h_to_4h__weight", "transformer.h.35.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__35__mlp__dense_h_to_4h__bias", "transformer.h.35.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__35__mlp__dense_4h_to_h__weight", "transformer.h.35.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__35__mlp__dense_4h_to_h__bias", "transformer.h.36.input_layernorm.weight": "bloomforcausallm__transformer__h__36__input_layernorm__weight", "transformer.h.36.input_layernorm.bias": "bloomforcausallm__transformer__h__36__input_layernorm__bias", "transformer.h.36.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__36__self_attention__query_key_value__weight", "transformer.h.36.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__36__self_attention__query_key_value__bias", "transformer.h.36.self_attention.dense.weight": "bloomforcausallm__transformer__h__36__self_attention__dense__weight", "transformer.h.36.self_attention.dense.bias": "bloomforcausallm__transformer__h__36__self_attention__dense__bias", "transformer.h.36.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__36__post_attention_layernorm__weight", "transformer.h.36.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__36__post_attention_layernorm__bias", "transformer.h.36.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__36__mlp__dense_h_to_4h__weight", "transformer.h.36.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__36__mlp__dense_h_to_4h__bias", "transformer.h.36.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__36__mlp__dense_4h_to_h__weight", "transformer.h.36.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__36__mlp__dense_4h_to_h__bias", "transformer.h.37.input_layernorm.weight": "bloomforcausallm__transformer__h__37__input_layernorm__weight", "transformer.h.37.input_layernorm.bias": "bloomforcausallm__transformer__h__37__input_layernorm__bias", "transformer.h.37.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__37__self_attention__query_key_value__weight", "transformer.h.37.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__37__self_attention__query_key_value__bias", "transformer.h.37.self_attention.dense.weight": "bloomforcausallm__transformer__h__37__self_attention__dense__weight", "transformer.h.37.self_attention.dense.bias": "bloomforcausallm__transformer__h__37__self_attention__dense__bias", "transformer.h.37.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__37__post_attention_layernorm__weight", "transformer.h.37.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__37__post_attention_layernorm__bias", "transformer.h.37.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__37__mlp__dense_h_to_4h__weight", "transformer.h.37.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__37__mlp__dense_h_to_4h__bias", "transformer.h.37.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__37__mlp__dense_4h_to_h__weight", "transformer.h.37.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__37__mlp__dense_4h_to_h__bias", "transformer.h.38.input_layernorm.weight": "bloomforcausallm__transformer__h__38__input_layernorm__weight", "transformer.h.38.input_layernorm.bias": "bloomforcausallm__transformer__h__38__input_layernorm__bias", "transformer.h.38.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__38__self_attention__query_key_value__weight", "transformer.h.38.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__38__self_attention__query_key_value__bias", "transformer.h.38.self_attention.dense.weight": "bloomforcausallm__transformer__h__38__self_attention__dense__weight", "transformer.h.38.self_attention.dense.bias": "bloomforcausallm__transformer__h__38__self_attention__dense__bias", "transformer.h.38.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__38__post_attention_layernorm__weight", "transformer.h.38.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__38__post_attention_layernorm__bias", "transformer.h.38.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__38__mlp__dense_h_to_4h__weight", "transformer.h.38.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__38__mlp__dense_h_to_4h__bias", "transformer.h.38.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__38__mlp__dense_4h_to_h__weight", "transformer.h.38.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__38__mlp__dense_4h_to_h__bias", "transformer.h.39.input_layernorm.weight": "bloomforcausallm__transformer__h__39__input_layernorm__weight", "transformer.h.39.input_layernorm.bias": "bloomforcausallm__transformer__h__39__input_layernorm__bias", "transformer.h.39.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__39__self_attention__query_key_value__weight", "transformer.h.39.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__39__self_attention__query_key_value__bias", "transformer.h.39.self_attention.dense.weight": "bloomforcausallm__transformer__h__39__self_attention__dense__weight", "transformer.h.39.self_attention.dense.bias": "bloomforcausallm__transformer__h__39__self_attention__dense__bias", "transformer.h.39.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__39__post_attention_layernorm__weight", "transformer.h.39.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__39__post_attention_layernorm__bias", "transformer.h.39.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__39__mlp__dense_h_to_4h__weight", "transformer.h.39.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__39__mlp__dense_h_to_4h__bias", "transformer.h.39.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__39__mlp__dense_4h_to_h__weight", "transformer.h.39.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__39__mlp__dense_4h_to_h__bias", "transformer.h.40.input_layernorm.weight": "bloomforcausallm__transformer__h__40__input_layernorm__weight", "transformer.h.40.input_layernorm.bias": "bloomforcausallm__transformer__h__40__input_layernorm__bias", "transformer.h.40.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__40__self_attention__query_key_value__weight", "transformer.h.40.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__40__self_attention__query_key_value__bias", "transformer.h.40.self_attention.dense.weight": "bloomforcausallm__transformer__h__40__self_attention__dense__weight", "transformer.h.40.self_attention.dense.bias": "bloomforcausallm__transformer__h__40__self_attention__dense__bias", "transformer.h.40.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__40__post_attention_layernorm__weight", "transformer.h.40.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__40__post_attention_layernorm__bias", "transformer.h.40.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__40__mlp__dense_h_to_4h__weight", "transformer.h.40.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__40__mlp__dense_h_to_4h__bias", "transformer.h.40.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__40__mlp__dense_4h_to_h__weight", "transformer.h.40.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__40__mlp__dense_4h_to_h__bias", "transformer.h.41.input_layernorm.weight": "bloomforcausallm__transformer__h__41__input_layernorm__weight", "transformer.h.41.input_layernorm.bias": "bloomforcausallm__transformer__h__41__input_layernorm__bias", "transformer.h.41.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__41__self_attention__query_key_value__weight", "transformer.h.41.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__41__self_attention__query_key_value__bias", "transformer.h.41.self_attention.dense.weight": "bloomforcausallm__transformer__h__41__self_attention__dense__weight", "transformer.h.41.self_attention.dense.bias": "bloomforcausallm__transformer__h__41__self_attention__dense__bias", "transformer.h.41.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__41__post_attention_layernorm__weight", "transformer.h.41.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__41__post_attention_layernorm__bias", "transformer.h.41.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__41__mlp__dense_h_to_4h__weight", "transformer.h.41.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__41__mlp__dense_h_to_4h__bias", "transformer.h.41.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__41__mlp__dense_4h_to_h__weight", "transformer.h.41.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__41__mlp__dense_4h_to_h__bias", "transformer.h.42.input_layernorm.weight": "bloomforcausallm__transformer__h__42__input_layernorm__weight", "transformer.h.42.input_layernorm.bias": "bloomforcausallm__transformer__h__42__input_layernorm__bias", "transformer.h.42.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__42__self_attention__query_key_value__weight", "transformer.h.42.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__42__self_attention__query_key_value__bias", "transformer.h.42.self_attention.dense.weight": "bloomforcausallm__transformer__h__42__self_attention__dense__weight", "transformer.h.42.self_attention.dense.bias": "bloomforcausallm__transformer__h__42__self_attention__dense__bias", "transformer.h.42.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__42__post_attention_layernorm__weight", "transformer.h.42.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__42__post_attention_layernorm__bias", "transformer.h.42.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__42__mlp__dense_h_to_4h__weight", "transformer.h.42.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__42__mlp__dense_h_to_4h__bias", "transformer.h.42.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__42__mlp__dense_4h_to_h__weight", "transformer.h.42.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__42__mlp__dense_4h_to_h__bias", "transformer.h.43.input_layernorm.weight": "bloomforcausallm__transformer__h__43__input_layernorm__weight", "transformer.h.43.input_layernorm.bias": "bloomforcausallm__transformer__h__43__input_layernorm__bias", "transformer.h.43.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__43__self_attention__query_key_value__weight", "transformer.h.43.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__43__self_attention__query_key_value__bias", "transformer.h.43.self_attention.dense.weight": "bloomforcausallm__transformer__h__43__self_attention__dense__weight", "transformer.h.43.self_attention.dense.bias": "bloomforcausallm__transformer__h__43__self_attention__dense__bias", "transformer.h.43.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__43__post_attention_layernorm__weight", "transformer.h.43.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__43__post_attention_layernorm__bias", "transformer.h.43.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__43__mlp__dense_h_to_4h__weight", "transformer.h.43.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__43__mlp__dense_h_to_4h__bias", "transformer.h.43.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__43__mlp__dense_4h_to_h__weight", "transformer.h.43.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__43__mlp__dense_4h_to_h__bias", "transformer.h.44.input_layernorm.weight": "bloomforcausallm__transformer__h__44__input_layernorm__weight", "transformer.h.44.input_layernorm.bias": "bloomforcausallm__transformer__h__44__input_layernorm__bias", "transformer.h.44.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__44__self_attention__query_key_value__weight", "transformer.h.44.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__44__self_attention__query_key_value__bias", "transformer.h.44.self_attention.dense.weight": "bloomforcausallm__transformer__h__44__self_attention__dense__weight", "transformer.h.44.self_attention.dense.bias": "bloomforcausallm__transformer__h__44__self_attention__dense__bias", "transformer.h.44.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__44__post_attention_layernorm__weight", "transformer.h.44.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__44__post_attention_layernorm__bias", "transformer.h.44.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__44__mlp__dense_h_to_4h__weight", "transformer.h.44.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__44__mlp__dense_h_to_4h__bias", "transformer.h.44.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__44__mlp__dense_4h_to_h__weight", "transformer.h.44.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__44__mlp__dense_4h_to_h__bias", "transformer.h.45.input_layernorm.weight": "bloomforcausallm__transformer__h__45__input_layernorm__weight", "transformer.h.45.input_layernorm.bias": "bloomforcausallm__transformer__h__45__input_layernorm__bias", "transformer.h.45.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__45__self_attention__query_key_value__weight", "transformer.h.45.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__45__self_attention__query_key_value__bias", "transformer.h.45.self_attention.dense.weight": "bloomforcausallm__transformer__h__45__self_attention__dense__weight", "transformer.h.45.self_attention.dense.bias": "bloomforcausallm__transformer__h__45__self_attention__dense__bias", "transformer.h.45.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__45__post_attention_layernorm__weight", "transformer.h.45.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__45__post_attention_layernorm__bias", "transformer.h.45.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__45__mlp__dense_h_to_4h__weight", "transformer.h.45.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__45__mlp__dense_h_to_4h__bias", "transformer.h.45.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__45__mlp__dense_4h_to_h__weight", "transformer.h.45.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__45__mlp__dense_4h_to_h__bias", "transformer.h.46.input_layernorm.weight": "bloomforcausallm__transformer__h__46__input_layernorm__weight", "transformer.h.46.input_layernorm.bias": "bloomforcausallm__transformer__h__46__input_layernorm__bias", "transformer.h.46.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__46__self_attention__query_key_value__weight", "transformer.h.46.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__46__self_attention__query_key_value__bias", "transformer.h.46.self_attention.dense.weight": "bloomforcausallm__transformer__h__46__self_attention__dense__weight", "transformer.h.46.self_attention.dense.bias": "bloomforcausallm__transformer__h__46__self_attention__dense__bias", "transformer.h.46.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__46__post_attention_layernorm__weight", "transformer.h.46.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__46__post_attention_layernorm__bias", "transformer.h.46.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__46__mlp__dense_h_to_4h__weight", "transformer.h.46.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__46__mlp__dense_h_to_4h__bias", "transformer.h.46.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__46__mlp__dense_4h_to_h__weight", "transformer.h.46.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__46__mlp__dense_4h_to_h__bias", "transformer.h.47.input_layernorm.weight": "bloomforcausallm__transformer__h__47__input_layernorm__weight", "transformer.h.47.input_layernorm.bias": "bloomforcausallm__transformer__h__47__input_layernorm__bias", "transformer.h.47.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__47__self_attention__query_key_value__weight", "transformer.h.47.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__47__self_attention__query_key_value__bias", "transformer.h.47.self_attention.dense.weight": "bloomforcausallm__transformer__h__47__self_attention__dense__weight", "transformer.h.47.self_attention.dense.bias": "bloomforcausallm__transformer__h__47__self_attention__dense__bias", "transformer.h.47.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__47__post_attention_layernorm__weight", "transformer.h.47.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__47__post_attention_layernorm__bias", "transformer.h.47.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__47__mlp__dense_h_to_4h__weight", "transformer.h.47.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__47__mlp__dense_h_to_4h__bias", "transformer.h.47.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__47__mlp__dense_4h_to_h__weight", "transformer.h.47.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__47__mlp__dense_4h_to_h__bias", "transformer.h.48.input_layernorm.weight": "bloomforcausallm__transformer__h__48__input_layernorm__weight", "transformer.h.48.input_layernorm.bias": "bloomforcausallm__transformer__h__48__input_layernorm__bias", "transformer.h.48.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__48__self_attention__query_key_value__weight", "transformer.h.48.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__48__self_attention__query_key_value__bias", "transformer.h.48.self_attention.dense.weight": "bloomforcausallm__transformer__h__48__self_attention__dense__weight", "transformer.h.48.self_attention.dense.bias": "bloomforcausallm__transformer__h__48__self_attention__dense__bias", "transformer.h.48.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__48__post_attention_layernorm__weight", "transformer.h.48.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__48__post_attention_layernorm__bias", "transformer.h.48.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__48__mlp__dense_h_to_4h__weight", "transformer.h.48.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__48__mlp__dense_h_to_4h__bias", "transformer.h.48.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__48__mlp__dense_4h_to_h__weight", "transformer.h.48.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__48__mlp__dense_4h_to_h__bias", "transformer.h.49.input_layernorm.weight": "bloomforcausallm__transformer__h__49__input_layernorm__weight", "transformer.h.49.input_layernorm.bias": "bloomforcausallm__transformer__h__49__input_layernorm__bias", "transformer.h.49.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__49__self_attention__query_key_value__weight", "transformer.h.49.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__49__self_attention__query_key_value__bias", "transformer.h.49.self_attention.dense.weight": "bloomforcausallm__transformer__h__49__self_attention__dense__weight", "transformer.h.49.self_attention.dense.bias": "bloomforcausallm__transformer__h__49__self_attention__dense__bias", "transformer.h.49.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__49__post_attention_layernorm__weight", "transformer.h.49.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__49__post_attention_layernorm__bias", "transformer.h.49.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__49__mlp__dense_h_to_4h__weight", "transformer.h.49.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__49__mlp__dense_h_to_4h__bias", "transformer.h.49.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__49__mlp__dense_4h_to_h__weight", "transformer.h.49.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__49__mlp__dense_4h_to_h__bias", "transformer.h.50.input_layernorm.weight": "bloomforcausallm__transformer__h__50__input_layernorm__weight", "transformer.h.50.input_layernorm.bias": "bloomforcausallm__transformer__h__50__input_layernorm__bias", "transformer.h.50.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__50__self_attention__query_key_value__weight", "transformer.h.50.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__50__self_attention__query_key_value__bias", "transformer.h.50.self_attention.dense.weight": "bloomforcausallm__transformer__h__50__self_attention__dense__weight", "transformer.h.50.self_attention.dense.bias": "bloomforcausallm__transformer__h__50__self_attention__dense__bias", "transformer.h.50.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__50__post_attention_layernorm__weight", "transformer.h.50.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__50__post_attention_layernorm__bias", "transformer.h.50.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__50__mlp__dense_h_to_4h__weight", "transformer.h.50.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__50__mlp__dense_h_to_4h__bias", "transformer.h.50.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__50__mlp__dense_4h_to_h__weight", "transformer.h.50.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__50__mlp__dense_4h_to_h__bias", "transformer.h.51.input_layernorm.weight": "bloomforcausallm__transformer__h__51__input_layernorm__weight", "transformer.h.51.input_layernorm.bias": "bloomforcausallm__transformer__h__51__input_layernorm__bias", "transformer.h.51.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__51__self_attention__query_key_value__weight", "transformer.h.51.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__51__self_attention__query_key_value__bias", "transformer.h.51.self_attention.dense.weight": "bloomforcausallm__transformer__h__51__self_attention__dense__weight", "transformer.h.51.self_attention.dense.bias": "bloomforcausallm__transformer__h__51__self_attention__dense__bias", "transformer.h.51.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__51__post_attention_layernorm__weight", "transformer.h.51.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__51__post_attention_layernorm__bias", "transformer.h.51.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__51__mlp__dense_h_to_4h__weight", "transformer.h.51.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__51__mlp__dense_h_to_4h__bias", "transformer.h.51.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__51__mlp__dense_4h_to_h__weight", "transformer.h.51.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__51__mlp__dense_4h_to_h__bias", "transformer.h.52.input_layernorm.weight": "bloomforcausallm__transformer__h__52__input_layernorm__weight", "transformer.h.52.input_layernorm.bias": "bloomforcausallm__transformer__h__52__input_layernorm__bias", "transformer.h.52.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__52__self_attention__query_key_value__weight", "transformer.h.52.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__52__self_attention__query_key_value__bias", "transformer.h.52.self_attention.dense.weight": "bloomforcausallm__transformer__h__52__self_attention__dense__weight", "transformer.h.52.self_attention.dense.bias": "bloomforcausallm__transformer__h__52__self_attention__dense__bias", "transformer.h.52.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__52__post_attention_layernorm__weight", "transformer.h.52.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__52__post_attention_layernorm__bias", "transformer.h.52.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__52__mlp__dense_h_to_4h__weight", "transformer.h.52.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__52__mlp__dense_h_to_4h__bias", "transformer.h.52.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__52__mlp__dense_4h_to_h__weight", "transformer.h.52.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__52__mlp__dense_4h_to_h__bias", "transformer.h.53.input_layernorm.weight": "bloomforcausallm__transformer__h__53__input_layernorm__weight", "transformer.h.53.input_layernorm.bias": "bloomforcausallm__transformer__h__53__input_layernorm__bias", "transformer.h.53.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__53__self_attention__query_key_value__weight", "transformer.h.53.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__53__self_attention__query_key_value__bias", "transformer.h.53.self_attention.dense.weight": "bloomforcausallm__transformer__h__53__self_attention__dense__weight", "transformer.h.53.self_attention.dense.bias": "bloomforcausallm__transformer__h__53__self_attention__dense__bias", "transformer.h.53.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__53__post_attention_layernorm__weight", "transformer.h.53.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__53__post_attention_layernorm__bias", "transformer.h.53.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__53__mlp__dense_h_to_4h__weight", "transformer.h.53.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__53__mlp__dense_h_to_4h__bias", "transformer.h.53.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__53__mlp__dense_4h_to_h__weight", "transformer.h.53.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__53__mlp__dense_4h_to_h__bias", "transformer.h.54.input_layernorm.weight": "bloomforcausallm__transformer__h__54__input_layernorm__weight", "transformer.h.54.input_layernorm.bias": "bloomforcausallm__transformer__h__54__input_layernorm__bias", "transformer.h.54.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__54__self_attention__query_key_value__weight", "transformer.h.54.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__54__self_attention__query_key_value__bias", "transformer.h.54.self_attention.dense.weight": "bloomforcausallm__transformer__h__54__self_attention__dense__weight", "transformer.h.54.self_attention.dense.bias": "bloomforcausallm__transformer__h__54__self_attention__dense__bias", "transformer.h.54.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__54__post_attention_layernorm__weight", "transformer.h.54.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__54__post_attention_layernorm__bias", "transformer.h.54.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__54__mlp__dense_h_to_4h__weight", "transformer.h.54.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__54__mlp__dense_h_to_4h__bias", "transformer.h.54.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__54__mlp__dense_4h_to_h__weight", "transformer.h.54.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__54__mlp__dense_4h_to_h__bias", "transformer.h.55.input_layernorm.weight": "bloomforcausallm__transformer__h__55__input_layernorm__weight", "transformer.h.55.input_layernorm.bias": "bloomforcausallm__transformer__h__55__input_layernorm__bias", "transformer.h.55.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__55__self_attention__query_key_value__weight", "transformer.h.55.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__55__self_attention__query_key_value__bias", "transformer.h.55.self_attention.dense.weight": "bloomforcausallm__transformer__h__55__self_attention__dense__weight", "transformer.h.55.self_attention.dense.bias": "bloomforcausallm__transformer__h__55__self_attention__dense__bias", "transformer.h.55.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__55__post_attention_layernorm__weight", "transformer.h.55.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__55__post_attention_layernorm__bias", "transformer.h.55.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__55__mlp__dense_h_to_4h__weight", "transformer.h.55.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__55__mlp__dense_h_to_4h__bias", "transformer.h.55.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__55__mlp__dense_4h_to_h__weight", "transformer.h.55.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__55__mlp__dense_4h_to_h__bias", "transformer.h.56.input_layernorm.weight": "bloomforcausallm__transformer__h__56__input_layernorm__weight", "transformer.h.56.input_layernorm.bias": "bloomforcausallm__transformer__h__56__input_layernorm__bias", "transformer.h.56.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__56__self_attention__query_key_value__weight", "transformer.h.56.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__56__self_attention__query_key_value__bias", "transformer.h.56.self_attention.dense.weight": "bloomforcausallm__transformer__h__56__self_attention__dense__weight", "transformer.h.56.self_attention.dense.bias": "bloomforcausallm__transformer__h__56__self_attention__dense__bias", "transformer.h.56.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__56__post_attention_layernorm__weight", "transformer.h.56.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__56__post_attention_layernorm__bias", "transformer.h.56.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__56__mlp__dense_h_to_4h__weight", "transformer.h.56.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__56__mlp__dense_h_to_4h__bias", "transformer.h.56.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__56__mlp__dense_4h_to_h__weight", "transformer.h.56.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__56__mlp__dense_4h_to_h__bias", "transformer.h.57.input_layernorm.weight": "bloomforcausallm__transformer__h__57__input_layernorm__weight", "transformer.h.57.input_layernorm.bias": "bloomforcausallm__transformer__h__57__input_layernorm__bias", "transformer.h.57.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__57__self_attention__query_key_value__weight", "transformer.h.57.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__57__self_attention__query_key_value__bias", "transformer.h.57.self_attention.dense.weight": "bloomforcausallm__transformer__h__57__self_attention__dense__weight", "transformer.h.57.self_attention.dense.bias": "bloomforcausallm__transformer__h__57__self_attention__dense__bias", "transformer.h.57.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__57__post_attention_layernorm__weight", "transformer.h.57.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__57__post_attention_layernorm__bias", "transformer.h.57.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__57__mlp__dense_h_to_4h__weight", "transformer.h.57.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__57__mlp__dense_h_to_4h__bias", "transformer.h.57.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__57__mlp__dense_4h_to_h__weight", "transformer.h.57.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__57__mlp__dense_4h_to_h__bias", "transformer.h.58.input_layernorm.weight": "bloomforcausallm__transformer__h__58__input_layernorm__weight", "transformer.h.58.input_layernorm.bias": "bloomforcausallm__transformer__h__58__input_layernorm__bias", "transformer.h.58.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__58__self_attention__query_key_value__weight", "transformer.h.58.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__58__self_attention__query_key_value__bias", "transformer.h.58.self_attention.dense.weight": "bloomforcausallm__transformer__h__58__self_attention__dense__weight", "transformer.h.58.self_attention.dense.bias": "bloomforcausallm__transformer__h__58__self_attention__dense__bias", "transformer.h.58.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__58__post_attention_layernorm__weight", "transformer.h.58.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__58__post_attention_layernorm__bias", "transformer.h.58.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__58__mlp__dense_h_to_4h__weight", "transformer.h.58.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__58__mlp__dense_h_to_4h__bias", "transformer.h.58.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__58__mlp__dense_4h_to_h__weight", "transformer.h.58.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__58__mlp__dense_4h_to_h__bias", "transformer.h.59.input_layernorm.weight": "bloomforcausallm__transformer__h__59__input_layernorm__weight", "transformer.h.59.input_layernorm.bias": "bloomforcausallm__transformer__h__59__input_layernorm__bias", "transformer.h.59.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__59__self_attention__query_key_value__weight", "transformer.h.59.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__59__self_attention__query_key_value__bias", "transformer.h.59.self_attention.dense.weight": "bloomforcausallm__transformer__h__59__self_attention__dense__weight", "transformer.h.59.self_attention.dense.bias": "bloomforcausallm__transformer__h__59__self_attention__dense__bias", "transformer.h.59.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__59__post_attention_layernorm__weight", "transformer.h.59.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__59__post_attention_layernorm__bias", "transformer.h.59.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__59__mlp__dense_h_to_4h__weight", "transformer.h.59.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__59__mlp__dense_h_to_4h__bias", "transformer.h.59.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__59__mlp__dense_4h_to_h__weight", "transformer.h.59.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__59__mlp__dense_4h_to_h__bias", "transformer.h.60.input_layernorm.weight": "bloomforcausallm__transformer__h__60__input_layernorm__weight", "transformer.h.60.input_layernorm.bias": "bloomforcausallm__transformer__h__60__input_layernorm__bias", "transformer.h.60.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__60__self_attention__query_key_value__weight", "transformer.h.60.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__60__self_attention__query_key_value__bias", "transformer.h.60.self_attention.dense.weight": "bloomforcausallm__transformer__h__60__self_attention__dense__weight", "transformer.h.60.self_attention.dense.bias": "bloomforcausallm__transformer__h__60__self_attention__dense__bias", "transformer.h.60.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__60__post_attention_layernorm__weight", "transformer.h.60.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__60__post_attention_layernorm__bias", "transformer.h.60.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__60__mlp__dense_h_to_4h__weight", "transformer.h.60.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__60__mlp__dense_h_to_4h__bias", "transformer.h.60.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__60__mlp__dense_4h_to_h__weight", "transformer.h.60.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__60__mlp__dense_4h_to_h__bias", "transformer.h.61.input_layernorm.weight": "bloomforcausallm__transformer__h__61__input_layernorm__weight", "transformer.h.61.input_layernorm.bias": "bloomforcausallm__transformer__h__61__input_layernorm__bias", "transformer.h.61.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__61__self_attention__query_key_value__weight", "transformer.h.61.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__61__self_attention__query_key_value__bias", "transformer.h.61.self_attention.dense.weight": "bloomforcausallm__transformer__h__61__self_attention__dense__weight", "transformer.h.61.self_attention.dense.bias": "bloomforcausallm__transformer__h__61__self_attention__dense__bias", "transformer.h.61.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__61__post_attention_layernorm__weight", "transformer.h.61.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__61__post_attention_layernorm__bias", "transformer.h.61.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__61__mlp__dense_h_to_4h__weight", "transformer.h.61.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__61__mlp__dense_h_to_4h__bias", "transformer.h.61.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__61__mlp__dense_4h_to_h__weight", "transformer.h.61.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__61__mlp__dense_4h_to_h__bias", "transformer.h.62.input_layernorm.weight": "bloomforcausallm__transformer__h__62__input_layernorm__weight", "transformer.h.62.input_layernorm.bias": "bloomforcausallm__transformer__h__62__input_layernorm__bias", "transformer.h.62.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__62__self_attention__query_key_value__weight", "transformer.h.62.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__62__self_attention__query_key_value__bias", "transformer.h.62.self_attention.dense.weight": "bloomforcausallm__transformer__h__62__self_attention__dense__weight", "transformer.h.62.self_attention.dense.bias": "bloomforcausallm__transformer__h__62__self_attention__dense__bias", "transformer.h.62.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__62__post_attention_layernorm__weight", "transformer.h.62.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__62__post_attention_layernorm__bias", "transformer.h.62.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__62__mlp__dense_h_to_4h__weight", "transformer.h.62.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__62__mlp__dense_h_to_4h__bias", "transformer.h.62.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__62__mlp__dense_4h_to_h__weight", "transformer.h.62.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__62__mlp__dense_4h_to_h__bias", "transformer.h.63.input_layernorm.weight": "bloomforcausallm__transformer__h__63__input_layernorm__weight", "transformer.h.63.input_layernorm.bias": "bloomforcausallm__transformer__h__63__input_layernorm__bias", "transformer.h.63.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__63__self_attention__query_key_value__weight", "transformer.h.63.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__63__self_attention__query_key_value__bias", "transformer.h.63.self_attention.dense.weight": "bloomforcausallm__transformer__h__63__self_attention__dense__weight", "transformer.h.63.self_attention.dense.bias": "bloomforcausallm__transformer__h__63__self_attention__dense__bias", "transformer.h.63.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__63__post_attention_layernorm__weight", "transformer.h.63.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__63__post_attention_layernorm__bias", "transformer.h.63.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__63__mlp__dense_h_to_4h__weight", "transformer.h.63.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__63__mlp__dense_h_to_4h__bias", "transformer.h.63.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__63__mlp__dense_4h_to_h__weight", "transformer.h.63.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__63__mlp__dense_4h_to_h__bias", "transformer.h.64.input_layernorm.weight": "bloomforcausallm__transformer__h__64__input_layernorm__weight", "transformer.h.64.input_layernorm.bias": "bloomforcausallm__transformer__h__64__input_layernorm__bias", "transformer.h.64.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__64__self_attention__query_key_value__weight", "transformer.h.64.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__64__self_attention__query_key_value__bias", "transformer.h.64.self_attention.dense.weight": "bloomforcausallm__transformer__h__64__self_attention__dense__weight", "transformer.h.64.self_attention.dense.bias": "bloomforcausallm__transformer__h__64__self_attention__dense__bias", "transformer.h.64.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__64__post_attention_layernorm__weight", "transformer.h.64.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__64__post_attention_layernorm__bias", "transformer.h.64.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__64__mlp__dense_h_to_4h__weight", "transformer.h.64.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__64__mlp__dense_h_to_4h__bias", "transformer.h.64.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__64__mlp__dense_4h_to_h__weight", "transformer.h.64.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__64__mlp__dense_4h_to_h__bias", "transformer.h.65.input_layernorm.weight": "bloomforcausallm__transformer__h__65__input_layernorm__weight", "transformer.h.65.input_layernorm.bias": "bloomforcausallm__transformer__h__65__input_layernorm__bias", "transformer.h.65.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__65__self_attention__query_key_value__weight", "transformer.h.65.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__65__self_attention__query_key_value__bias", "transformer.h.65.self_attention.dense.weight": "bloomforcausallm__transformer__h__65__self_attention__dense__weight", "transformer.h.65.self_attention.dense.bias": "bloomforcausallm__transformer__h__65__self_attention__dense__bias", "transformer.h.65.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__65__post_attention_layernorm__weight", "transformer.h.65.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__65__post_attention_layernorm__bias", "transformer.h.65.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__65__mlp__dense_h_to_4h__weight", "transformer.h.65.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__65__mlp__dense_h_to_4h__bias", "transformer.h.65.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__65__mlp__dense_4h_to_h__weight", "transformer.h.65.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__65__mlp__dense_4h_to_h__bias", "transformer.h.66.input_layernorm.weight": "bloomforcausallm__transformer__h__66__input_layernorm__weight", "transformer.h.66.input_layernorm.bias": "bloomforcausallm__transformer__h__66__input_layernorm__bias", "transformer.h.66.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__66__self_attention__query_key_value__weight", "transformer.h.66.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__66__self_attention__query_key_value__bias", "transformer.h.66.self_attention.dense.weight": "bloomforcausallm__transformer__h__66__self_attention__dense__weight", "transformer.h.66.self_attention.dense.bias": "bloomforcausallm__transformer__h__66__self_attention__dense__bias", "transformer.h.66.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__66__post_attention_layernorm__weight", "transformer.h.66.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__66__post_attention_layernorm__bias", "transformer.h.66.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__66__mlp__dense_h_to_4h__weight", "transformer.h.66.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__66__mlp__dense_h_to_4h__bias", "transformer.h.66.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__66__mlp__dense_4h_to_h__weight", "transformer.h.66.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__66__mlp__dense_4h_to_h__bias", "transformer.h.67.input_layernorm.weight": "bloomforcausallm__transformer__h__67__input_layernorm__weight", "transformer.h.67.input_layernorm.bias": "bloomforcausallm__transformer__h__67__input_layernorm__bias", "transformer.h.67.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__67__self_attention__query_key_value__weight", "transformer.h.67.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__67__self_attention__query_key_value__bias", "transformer.h.67.self_attention.dense.weight": "bloomforcausallm__transformer__h__67__self_attention__dense__weight", "transformer.h.67.self_attention.dense.bias": "bloomforcausallm__transformer__h__67__self_attention__dense__bias", "transformer.h.67.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__67__post_attention_layernorm__weight", "transformer.h.67.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__67__post_attention_layernorm__bias", "transformer.h.67.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__67__mlp__dense_h_to_4h__weight", "transformer.h.67.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__67__mlp__dense_h_to_4h__bias", "transformer.h.67.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__67__mlp__dense_4h_to_h__weight", "transformer.h.67.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__67__mlp__dense_4h_to_h__bias", "transformer.h.68.input_layernorm.weight": "bloomforcausallm__transformer__h__68__input_layernorm__weight", "transformer.h.68.input_layernorm.bias": "bloomforcausallm__transformer__h__68__input_layernorm__bias", "transformer.h.68.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__68__self_attention__query_key_value__weight", "transformer.h.68.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__68__self_attention__query_key_value__bias", "transformer.h.68.self_attention.dense.weight": "bloomforcausallm__transformer__h__68__self_attention__dense__weight", "transformer.h.68.self_attention.dense.bias": "bloomforcausallm__transformer__h__68__self_attention__dense__bias", "transformer.h.68.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__68__post_attention_layernorm__weight", "transformer.h.68.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__68__post_attention_layernorm__bias", "transformer.h.68.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__68__mlp__dense_h_to_4h__weight", "transformer.h.68.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__68__mlp__dense_h_to_4h__bias", "transformer.h.68.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__68__mlp__dense_4h_to_h__weight", "transformer.h.68.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__68__mlp__dense_4h_to_h__bias", "transformer.h.69.input_layernorm.weight": "bloomforcausallm__transformer__h__69__input_layernorm__weight", "transformer.h.69.input_layernorm.bias": "bloomforcausallm__transformer__h__69__input_layernorm__bias", "transformer.h.69.self_attention.query_key_value.weight": "bloomforcausallm__transformer__h__69__self_attention__query_key_value__weight", "transformer.h.69.self_attention.query_key_value.bias": "bloomforcausallm__transformer__h__69__self_attention__query_key_value__bias", "transformer.h.69.self_attention.dense.weight": "bloomforcausallm__transformer__h__69__self_attention__dense__weight", "transformer.h.69.self_attention.dense.bias": "bloomforcausallm__transformer__h__69__self_attention__dense__bias", "transformer.h.69.post_attention_layernorm.weight": "bloomforcausallm__transformer__h__69__post_attention_layernorm__weight", "transformer.h.69.post_attention_layernorm.bias": "bloomforcausallm__transformer__h__69__post_attention_layernorm__bias", "transformer.h.69.mlp.dense_h_to_4h.weight": "bloomforcausallm__transformer__h__69__mlp__dense_h_to_4h__weight", "transformer.h.69.mlp.dense_h_to_4h.bias": "bloomforcausallm__transformer__h__69__mlp__dense_h_to_4h__bias", "transformer.h.69.mlp.dense_4h_to_h.weight": "bloomforcausallm__transformer__h__69__mlp__dense_4h_to_h__weight", "transformer.h.69.mlp.dense_4h_to_h.bias": "bloomforcausallm__transformer__h__69__mlp__dense_4h_to_h__bias", "transformer.ln_f.weight": "bloomforcausallm__transformer__ln_f__weight", "transformer.ln_f.bias": "bloomforcausallm__transformer__ln_f__bias", "lm_head.weight": "bloomforcausallm__transformer__word_embeddings__weight"}
special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "pad_token": "<pad>",
5
+ "unk_token": "<unk>"
6
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f6efc66e73f1fd69da4f436e48befb519fdff3fe18910850c1d41bd862293a5
3
+ size 14500443
tokenizer_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "bos_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "model_max_length": 1000000000000000019884624838656,
6
+ "name_or_path": "bigscience/bloom",
7
+ "pad_token": "<pad>",
8
+ "padding_side": "left",
9
+ "special_tokens_map_file": null,
10
+ "tokenizer_class": "BloomTokenizer",
11
+ "unk_token": "<unk>"
12
+ }