Hugging Face's logo

oscorrea
/
tt-turbot-ai-07-03-24

Text Generation
Transformers
TensorBoard
Safetensors
mixtral
conversational
text-generation-inference
Inference Endpoints
Model card Files Metrics Community
oscorrea commited on
Commit
f3db725
·
verified ·
1 Parent(s): 6a14a79

Upload folder using huggingface_hub

Browse files
Files changed (34) hide show
  1. adapter_config.json +33 -0
  2. adapter_model.safetensors +3 -0
  3. all_results.json +14 -0
  4. config.json +30 -0
  5. eval_results.json +9 -0
  6. generation_config.json +6 -0
  7. model-00001-of-00019.safetensors +3 -0
  8. model-00002-of-00019.safetensors +3 -0
  9. model-00003-of-00019.safetensors +3 -0
  10. model-00004-of-00019.safetensors +3 -0
  11. model-00005-of-00019.safetensors +3 -0
  12. model-00006-of-00019.safetensors +3 -0
  13. model-00007-of-00019.safetensors +3 -0
  14. model-00008-of-00019.safetensors +3 -0
  15. model-00009-of-00019.safetensors +3 -0
  16. model-00010-of-00019.safetensors +3 -0
  17. model-00011-of-00019.safetensors +3 -0
  18. model-00012-of-00019.safetensors +3 -0
  19. model-00013-of-00019.safetensors +3 -0
  20. model-00014-of-00019.safetensors +3 -0
  21. model-00015-of-00019.safetensors +3 -0
  22. model-00016-of-00019.safetensors +3 -0
  23. model-00017-of-00019.safetensors +3 -0
  24. model-00018-of-00019.safetensors +3 -0
  25. model-00019-of-00019.safetensors +3 -0
  26. model.safetensors.index.json +1002 -0
  27. runs/Mar07_15-07-19_algo-1/events.out.tfevents.1709824095.algo-1.118.0 +3 -0
  28. runs/Mar07_15-07-19_algo-1/events.out.tfevents.1709854397.algo-1.118.1 +3 -0
  29. special_tokens_map.json +28 -0
  30. tokenizer.json +0 -0
  31. tokenizer.model +3 -0
  32. tokenizer_config.json +46 -0
  33. train_results.json +8 -0
  34. trainer_state.json +394 -0
adapter_config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "/tmp",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layers_pattern": null,
10
+ "layers_to_transform": null,
11
+ "loftq_config": {},
12
+ "lora_alpha": 16,
13
+ "lora_dropout": 0.0,
14
+ "megatron_config": null,
15
+ "megatron_core": "megatron.core",
16
+ "modules_to_save": null,
17
+ "peft_type": "LORA",
18
+ "r": 64,
19
+ "rank_pattern": {},
20
+ "revision": null,
21
+ "target_modules": [
22
+ "o_proj",
23
+ "gate",
24
+ "w1",
25
+ "k_proj",
26
+ "w2",
27
+ "v_proj",
28
+ "q_proj",
29
+ "w3"
30
+ ],
31
+ "task_type": "CAUSAL_LM",
32
+ "use_rslora": false
33
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ae067fffa2924996e6a1cc9d132a8649955dd709639ff447dbb44d1af620a70
3
+ size 1938079368
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_loss": 0.3297732174396515,
4
+ "eval_runtime": 1099.1337,
5
+ "eval_samples": 3578,
6
+ "eval_samples_per_second": 1.039,
7
+ "eval_steps_per_second": 0.13,
8
+ "perplexity": 1.390652716916515,
9
+ "train_loss": 0.5342465417401601,
10
+ "train_runtime": 29202.9895,
11
+ "train_samples": 14310,
12
+ "train_samples_per_second": 0.158,
13
+ "train_steps_per_second": 0.01
14
+ }
config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/tmp",
3
+ "architectures": [
4
+ "MixtralForCausalLM"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 1,
8
+ "eos_token_id": 2,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 4096,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 14336,
13
+ "max_position_embeddings": 32768,
14
+ "model_type": "mixtral",
15
+ "num_attention_heads": 32,
16
+ "num_experts_per_tok": 2,
17
+ "num_hidden_layers": 32,
18
+ "num_key_value_heads": 8,
19
+ "num_local_experts": 8,
20
+ "output_router_logits": false,
21
+ "rms_norm_eps": 1e-05,
22
+ "rope_theta": 1000000.0,
23
+ "router_aux_loss_coef": 0.02,
24
+ "sliding_window": 4096,
25
+ "tie_word_embeddings": false,
26
+ "torch_dtype": "float16",
27
+ "transformers_version": "4.38.1",
28
+ "use_cache": true,
29
+ "vocab_size": 32000
30
+ }
eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "eval_loss": 0.3297732174396515,
4
+ "eval_runtime": 1099.1337,
5
+ "eval_samples": 3578,
6
+ "eval_samples_per_second": 1.039,
7
+ "eval_steps_per_second": 0.13,
8
+ "perplexity": 1.390652716916515
9
+ }
generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 2,
5
+ "transformers_version": "4.38.1"
6
+ }
model-00001-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e1c663f1846a28816da616465175a95c2f6a99c1d412b2f74478cc95feab3abe
3
+ size 4892809528
model-00002-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c23afed23e839e7080bbe2538842b839605ac88d909c85ed1cf1c4de38b0897
3
+ size 4983003960
model-00003-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:375b5c059299443de6bc8809e0dcd38a3c93a2002cf3d561f570c79fb09027ae
3
+ size 4983003960
model-00004-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a39f6f934d553569b12d290e3f3cc5bf9447c7fc4905b599cf61521513b222f
3
+ size 4899035152
model-00005-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f554412d16455c0bc25905bb4f5272cd5b9bb65ceeaacb0a7fbbdaac7a6a583
3
+ size 4983003960
model-00006-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c266e0b6ea34f3bd0e5fa70e01bd4f9ebaf62f7bac90a1c3c02bb4d1307cb0e9
3
+ size 4983003960
model-00007-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e3642ae6855ac5a062949a247c47d089d762b48b2c346c33ea343d148a187aa
3
+ size 4899035200
model-00008-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b1116ea02c41dd5d0663222b290985ee608cf572d027e70f5a7ad2d13d8b767
3
+ size 4983004016
model-00009-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07daf7d1c3f9f1602be5add65e31b7d443ef1a5b8e6e59ce2d9e3f145984258d
3
+ size 4983004016
model-00010-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0c840a949523c52f54e165cf0af02a63aea3ffd2120081d1983d35c4b3389400
3
+ size 4899035200
model-00011-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0347956099e0f841635e6985cfcccb9b0d35d041c57c67b8d91d4cfcfc0927ec
3
+ size 4983004016
model-00012-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9201f3838c8142a388726f75e949f40b68a8284250b4550976f470035d81b28
3
+ size 4983004016
model-00013-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35754edb72fdaae4e028a85e79c5127e6f4206c5e8df883d80fa2769040c0cd0
3
+ size 4983004016
model-00014-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9a1423ed6487bfcb4963b6642366c01c6b3b7017f0895023083c5b0b0f9b8ad0
3
+ size 4899035200
model-00015-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb797b172c4532eda02ec1dd8f7e944d769672bbfb9f80d9c74cf5aae5871df8
3
+ size 4983004016
model-00016-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:52ccf30724b0cc23dfadb27c29166a2d5074399a2d423ed8af9a33198edb2860
3
+ size 4983004016
model-00017-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19fd00d38c109531c13ceba14e9fda2008cd43ef6cd3956f1cae647527f7a1aa
3
+ size 4899035200
model-00018-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:314cc18fdce317c59415b229a8342355cb704d3a41ba35f65d99abe5594213e1
3
+ size 4983004016
model-00019-of-00019.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af4a12dc6d9437679f4137fa56d49e3f31dfac0dc2f298b9636d265e777b378f
3
+ size 4221679048
model.safetensors.index.json ADDED
@@ -0,0 +1,1002 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 93405585408
4
+ },
5
+ "weight_map": {
6
+ "lm_head.weight": "model-00019-of-00019.safetensors",
7
+ "model.embed_tokens.weight": "model-00001-of-00019.safetensors",
8
+ "model.layers.0.block_sparse_moe.experts.0.w1.weight": "model-00001-of-00019.safetensors",
9
+ "model.layers.0.block_sparse_moe.experts.0.w2.weight": "model-00001-of-00019.safetensors",
10
+ "model.layers.0.block_sparse_moe.experts.0.w3.weight": "model-00001-of-00019.safetensors",
11
+ "model.layers.0.block_sparse_moe.experts.1.w1.weight": "model-00001-of-00019.safetensors",
12
+ "model.layers.0.block_sparse_moe.experts.1.w2.weight": "model-00001-of-00019.safetensors",
13
+ "model.layers.0.block_sparse_moe.experts.1.w3.weight": "model-00001-of-00019.safetensors",
14
+ "model.layers.0.block_sparse_moe.experts.2.w1.weight": "model-00001-of-00019.safetensors",
15
+ "model.layers.0.block_sparse_moe.experts.2.w2.weight": "model-00001-of-00019.safetensors",
16
+ "model.layers.0.block_sparse_moe.experts.2.w3.weight": "model-00001-of-00019.safetensors",
17
+ "model.layers.0.block_sparse_moe.experts.3.w1.weight": "model-00001-of-00019.safetensors",
18
+ "model.layers.0.block_sparse_moe.experts.3.w2.weight": "model-00001-of-00019.safetensors",
19
+ "model.layers.0.block_sparse_moe.experts.3.w3.weight": "model-00001-of-00019.safetensors",
20
+ "model.layers.0.block_sparse_moe.experts.4.w1.weight": "model-00001-of-00019.safetensors",
21
+ "model.layers.0.block_sparse_moe.experts.4.w2.weight": "model-00001-of-00019.safetensors",
22
+ "model.layers.0.block_sparse_moe.experts.4.w3.weight": "model-00001-of-00019.safetensors",
23
+ "model.layers.0.block_sparse_moe.experts.5.w1.weight": "model-00001-of-00019.safetensors",
24
+ "model.layers.0.block_sparse_moe.experts.5.w2.weight": "model-00001-of-00019.safetensors",
25
+ "model.layers.0.block_sparse_moe.experts.5.w3.weight": "model-00001-of-00019.safetensors",
26
+ "model.layers.0.block_sparse_moe.experts.6.w1.weight": "model-00001-of-00019.safetensors",
27
+ "model.layers.0.block_sparse_moe.experts.6.w2.weight": "model-00001-of-00019.safetensors",
28
+ "model.layers.0.block_sparse_moe.experts.6.w3.weight": "model-00001-of-00019.safetensors",
29
+ "model.layers.0.block_sparse_moe.experts.7.w1.weight": "model-00001-of-00019.safetensors",
30
+ "model.layers.0.block_sparse_moe.experts.7.w2.weight": "model-00001-of-00019.safetensors",
31
+ "model.layers.0.block_sparse_moe.experts.7.w3.weight": "model-00001-of-00019.safetensors",
32
+ "model.layers.0.block_sparse_moe.gate.weight": "model-00001-of-00019.safetensors",
33
+ "model.layers.0.input_layernorm.weight": "model-00001-of-00019.safetensors",
34
+ "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00019.safetensors",
35
+ "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00019.safetensors",
36
+ "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00019.safetensors",
37
+ "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00019.safetensors",
38
+ "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00019.safetensors",
39
+ "model.layers.1.block_sparse_moe.experts.0.w1.weight": "model-00001-of-00019.safetensors",
40
+ "model.layers.1.block_sparse_moe.experts.0.w2.weight": "model-00001-of-00019.safetensors",
41
+ "model.layers.1.block_sparse_moe.experts.0.w3.weight": "model-00001-of-00019.safetensors",
42
+ "model.layers.1.block_sparse_moe.experts.1.w1.weight": "model-00001-of-00019.safetensors",
43
+ "model.layers.1.block_sparse_moe.experts.1.w2.weight": "model-00001-of-00019.safetensors",
44
+ "model.layers.1.block_sparse_moe.experts.1.w3.weight": "model-00001-of-00019.safetensors",
45
+ "model.layers.1.block_sparse_moe.experts.2.w1.weight": "model-00001-of-00019.safetensors",
46
+ "model.layers.1.block_sparse_moe.experts.2.w2.weight": "model-00001-of-00019.safetensors",
47
+ "model.layers.1.block_sparse_moe.experts.2.w3.weight": "model-00001-of-00019.safetensors",
48
+ "model.layers.1.block_sparse_moe.experts.3.w1.weight": "model-00001-of-00019.safetensors",
49
+ "model.layers.1.block_sparse_moe.experts.3.w2.weight": "model-00001-of-00019.safetensors",
50
+ "model.layers.1.block_sparse_moe.experts.3.w3.weight": "model-00001-of-00019.safetensors",
51
+ "model.layers.1.block_sparse_moe.experts.4.w1.weight": "model-00001-of-00019.safetensors",
52
+ "model.layers.1.block_sparse_moe.experts.4.w2.weight": "model-00001-of-00019.safetensors",
53
+ "model.layers.1.block_sparse_moe.experts.4.w3.weight": "model-00002-of-00019.safetensors",
54
+ "model.layers.1.block_sparse_moe.experts.5.w1.weight": "model-00002-of-00019.safetensors",
55
+ "model.layers.1.block_sparse_moe.experts.5.w2.weight": "model-00002-of-00019.safetensors",
56
+ "model.layers.1.block_sparse_moe.experts.5.w3.weight": "model-00002-of-00019.safetensors",
57
+ "model.layers.1.block_sparse_moe.experts.6.w1.weight": "model-00002-of-00019.safetensors",
58
+ "model.layers.1.block_sparse_moe.experts.6.w2.weight": "model-00002-of-00019.safetensors",
59
+ "model.layers.1.block_sparse_moe.experts.6.w3.weight": "model-00002-of-00019.safetensors",
60
+ "model.layers.1.block_sparse_moe.experts.7.w1.weight": "model-00002-of-00019.safetensors",
61
+ "model.layers.1.block_sparse_moe.experts.7.w2.weight": "model-00002-of-00019.safetensors",
62
+ "model.layers.1.block_sparse_moe.experts.7.w3.weight": "model-00002-of-00019.safetensors",
63
+ "model.layers.1.block_sparse_moe.gate.weight": "model-00001-of-00019.safetensors",
64
+ "model.layers.1.input_layernorm.weight": "model-00002-of-00019.safetensors",
65
+ "model.layers.1.post_attention_layernorm.weight": "model-00002-of-00019.safetensors",
66
+ "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00019.safetensors",
67
+ "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00019.safetensors",
68
+ "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00019.safetensors",
69
+ "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00019.safetensors",
70
+ "model.layers.10.block_sparse_moe.experts.0.w1.weight": "model-00006-of-00019.safetensors",
71
+ "model.layers.10.block_sparse_moe.experts.0.w2.weight": "model-00006-of-00019.safetensors",
72
+ "model.layers.10.block_sparse_moe.experts.0.w3.weight": "model-00006-of-00019.safetensors",
73
+ "model.layers.10.block_sparse_moe.experts.1.w1.weight": "model-00007-of-00019.safetensors",
74
+ "model.layers.10.block_sparse_moe.experts.1.w2.weight": "model-00007-of-00019.safetensors",
75
+ "model.layers.10.block_sparse_moe.experts.1.w3.weight": "model-00007-of-00019.safetensors",
76
+ "model.layers.10.block_sparse_moe.experts.2.w1.weight": "model-00007-of-00019.safetensors",
77
+ "model.layers.10.block_sparse_moe.experts.2.w2.weight": "model-00007-of-00019.safetensors",
78
+ "model.layers.10.block_sparse_moe.experts.2.w3.weight": "model-00007-of-00019.safetensors",
79
+ "model.layers.10.block_sparse_moe.experts.3.w1.weight": "model-00007-of-00019.safetensors",
80
+ "model.layers.10.block_sparse_moe.experts.3.w2.weight": "model-00007-of-00019.safetensors",
81
+ "model.layers.10.block_sparse_moe.experts.3.w3.weight": "model-00007-of-00019.safetensors",
82
+ "model.layers.10.block_sparse_moe.experts.4.w1.weight": "model-00007-of-00019.safetensors",
83
+ "model.layers.10.block_sparse_moe.experts.4.w2.weight": "model-00007-of-00019.safetensors",
84
+ "model.layers.10.block_sparse_moe.experts.4.w3.weight": "model-00007-of-00019.safetensors",
85
+ "model.layers.10.block_sparse_moe.experts.5.w1.weight": "model-00007-of-00019.safetensors",
86
+ "model.layers.10.block_sparse_moe.experts.5.w2.weight": "model-00007-of-00019.safetensors",
87
+ "model.layers.10.block_sparse_moe.experts.5.w3.weight": "model-00007-of-00019.safetensors",
88
+ "model.layers.10.block_sparse_moe.experts.6.w1.weight": "model-00007-of-00019.safetensors",
89
+ "model.layers.10.block_sparse_moe.experts.6.w2.weight": "model-00007-of-00019.safetensors",
90
+ "model.layers.10.block_sparse_moe.experts.6.w3.weight": "model-00007-of-00019.safetensors",
91
+ "model.layers.10.block_sparse_moe.experts.7.w1.weight": "model-00007-of-00019.safetensors",
92
+ "model.layers.10.block_sparse_moe.experts.7.w2.weight": "model-00007-of-00019.safetensors",
93
+ "model.layers.10.block_sparse_moe.experts.7.w3.weight": "model-00007-of-00019.safetensors",
94
+ "model.layers.10.block_sparse_moe.gate.weight": "model-00006-of-00019.safetensors",
95
+ "model.layers.10.input_layernorm.weight": "model-00007-of-00019.safetensors",
96
+ "model.layers.10.post_attention_layernorm.weight": "model-00007-of-00019.safetensors",
97
+ "model.layers.10.self_attn.k_proj.weight": "model-00006-of-00019.safetensors",
98
+ "model.layers.10.self_attn.o_proj.weight": "model-00006-of-00019.safetensors",
99
+ "model.layers.10.self_attn.q_proj.weight": "model-00006-of-00019.safetensors",
100
+ "model.layers.10.self_attn.v_proj.weight": "model-00006-of-00019.safetensors",
101
+ "model.layers.11.block_sparse_moe.experts.0.w1.weight": "model-00007-of-00019.safetensors",
102
+ "model.layers.11.block_sparse_moe.experts.0.w2.weight": "model-00007-of-00019.safetensors",
103
+ "model.layers.11.block_sparse_moe.experts.0.w3.weight": "model-00007-of-00019.safetensors",
104
+ "model.layers.11.block_sparse_moe.experts.1.w1.weight": "model-00007-of-00019.safetensors",
105
+ "model.layers.11.block_sparse_moe.experts.1.w2.weight": "model-00007-of-00019.safetensors",
106
+ "model.layers.11.block_sparse_moe.experts.1.w3.weight": "model-00007-of-00019.safetensors",
107
+ "model.layers.11.block_sparse_moe.experts.2.w1.weight": "model-00007-of-00019.safetensors",
108
+ "model.layers.11.block_sparse_moe.experts.2.w2.weight": "model-00007-of-00019.safetensors",
109
+ "model.layers.11.block_sparse_moe.experts.2.w3.weight": "model-00007-of-00019.safetensors",
110
+ "model.layers.11.block_sparse_moe.experts.3.w1.weight": "model-00007-of-00019.safetensors",
111
+ "model.layers.11.block_sparse_moe.experts.3.w2.weight": "model-00007-of-00019.safetensors",
112
+ "model.layers.11.block_sparse_moe.experts.3.w3.weight": "model-00007-of-00019.safetensors",
113
+ "model.layers.11.block_sparse_moe.experts.4.w1.weight": "model-00007-of-00019.safetensors",
114
+ "model.layers.11.block_sparse_moe.experts.4.w2.weight": "model-00007-of-00019.safetensors",
115
+ "model.layers.11.block_sparse_moe.experts.4.w3.weight": "model-00007-of-00019.safetensors",
116
+ "model.layers.11.block_sparse_moe.experts.5.w1.weight": "model-00007-of-00019.safetensors",
117
+ "model.layers.11.block_sparse_moe.experts.5.w2.weight": "model-00007-of-00019.safetensors",
118
+ "model.layers.11.block_sparse_moe.experts.5.w3.weight": "model-00007-of-00019.safetensors",
119
+ "model.layers.11.block_sparse_moe.experts.6.w1.weight": "model-00007-of-00019.safetensors",
120
+ "model.layers.11.block_sparse_moe.experts.6.w2.weight": "model-00007-of-00019.safetensors",
121
+ "model.layers.11.block_sparse_moe.experts.6.w3.weight": "model-00008-of-00019.safetensors",
122
+ "model.layers.11.block_sparse_moe.experts.7.w1.weight": "model-00008-of-00019.safetensors",
123
+ "model.layers.11.block_sparse_moe.experts.7.w2.weight": "model-00008-of-00019.safetensors",
124
+ "model.layers.11.block_sparse_moe.experts.7.w3.weight": "model-00008-of-00019.safetensors",
125
+ "model.layers.11.block_sparse_moe.gate.weight": "model-00007-of-00019.safetensors",
126
+ "model.layers.11.input_layernorm.weight": "model-00008-of-00019.safetensors",
127
+ "model.layers.11.post_attention_layernorm.weight": "model-00008-of-00019.safetensors",
128
+ "model.layers.11.self_attn.k_proj.weight": "model-00007-of-00019.safetensors",
129
+ "model.layers.11.self_attn.o_proj.weight": "model-00007-of-00019.safetensors",
130
+ "model.layers.11.self_attn.q_proj.weight": "model-00007-of-00019.safetensors",
131
+ "model.layers.11.self_attn.v_proj.weight": "model-00007-of-00019.safetensors",
132
+ "model.layers.12.block_sparse_moe.experts.0.w1.weight": "model-00008-of-00019.safetensors",
133
+ "model.layers.12.block_sparse_moe.experts.0.w2.weight": "model-00008-of-00019.safetensors",
134
+ "model.layers.12.block_sparse_moe.experts.0.w3.weight": "model-00008-of-00019.safetensors",
135
+ "model.layers.12.block_sparse_moe.experts.1.w1.weight": "model-00008-of-00019.safetensors",
136
+ "model.layers.12.block_sparse_moe.experts.1.w2.weight": "model-00008-of-00019.safetensors",
137
+ "model.layers.12.block_sparse_moe.experts.1.w3.weight": "model-00008-of-00019.safetensors",
138
+ "model.layers.12.block_sparse_moe.experts.2.w1.weight": "model-00008-of-00019.safetensors",
139
+ "model.layers.12.block_sparse_moe.experts.2.w2.weight": "model-00008-of-00019.safetensors",
140
+ "model.layers.12.block_sparse_moe.experts.2.w3.weight": "model-00008-of-00019.safetensors",
141
+ "model.layers.12.block_sparse_moe.experts.3.w1.weight": "model-00008-of-00019.safetensors",
142
+ "model.layers.12.block_sparse_moe.experts.3.w2.weight": "model-00008-of-00019.safetensors",
143
+ "model.layers.12.block_sparse_moe.experts.3.w3.weight": "model-00008-of-00019.safetensors",
144
+ "model.layers.12.block_sparse_moe.experts.4.w1.weight": "model-00008-of-00019.safetensors",
145
+ "model.layers.12.block_sparse_moe.experts.4.w2.weight": "model-00008-of-00019.safetensors",
146
+ "model.layers.12.block_sparse_moe.experts.4.w3.weight": "model-00008-of-00019.safetensors",
147
+ "model.layers.12.block_sparse_moe.experts.5.w1.weight": "model-00008-of-00019.safetensors",
148
+ "model.layers.12.block_sparse_moe.experts.5.w2.weight": "model-00008-of-00019.safetensors",
149
+ "model.layers.12.block_sparse_moe.experts.5.w3.weight": "model-00008-of-00019.safetensors",
150
+ "model.layers.12.block_sparse_moe.experts.6.w1.weight": "model-00008-of-00019.safetensors",
151
+ "model.layers.12.block_sparse_moe.experts.6.w2.weight": "model-00008-of-00019.safetensors",
152
+ "model.layers.12.block_sparse_moe.experts.6.w3.weight": "model-00008-of-00019.safetensors",
153
+ "model.layers.12.block_sparse_moe.experts.7.w1.weight": "model-00008-of-00019.safetensors",
154
+ "model.layers.12.block_sparse_moe.experts.7.w2.weight": "model-00008-of-00019.safetensors",
155
+ "model.layers.12.block_sparse_moe.experts.7.w3.weight": "model-00008-of-00019.safetensors",
156
+ "model.layers.12.block_sparse_moe.gate.weight": "model-00008-of-00019.safetensors",
157
+ "model.layers.12.input_layernorm.weight": "model-00008-of-00019.safetensors",
158
+ "model.layers.12.post_attention_layernorm.weight": "model-00008-of-00019.safetensors",
159
+ "model.layers.12.self_attn.k_proj.weight": "model-00008-of-00019.safetensors",
160
+ "model.layers.12.self_attn.o_proj.weight": "model-00008-of-00019.safetensors",
161
+ "model.layers.12.self_attn.q_proj.weight": "model-00008-of-00019.safetensors",
162
+ "model.layers.12.self_attn.v_proj.weight": "model-00008-of-00019.safetensors",
163
+ "model.layers.13.block_sparse_moe.experts.0.w1.weight": "model-00008-of-00019.safetensors",
164
+ "model.layers.13.block_sparse_moe.experts.0.w2.weight": "model-00008-of-00019.safetensors",
165
+ "model.layers.13.block_sparse_moe.experts.0.w3.weight": "model-00008-of-00019.safetensors",
166
+ "model.layers.13.block_sparse_moe.experts.1.w1.weight": "model-00008-of-00019.safetensors",
167
+ "model.layers.13.block_sparse_moe.experts.1.w2.weight": "model-00008-of-00019.safetensors",
168
+ "model.layers.13.block_sparse_moe.experts.1.w3.weight": "model-00008-of-00019.safetensors",
169
+ "model.layers.13.block_sparse_moe.experts.2.w1.weight": "model-00008-of-00019.safetensors",
170
+ "model.layers.13.block_sparse_moe.experts.2.w2.weight": "model-00008-of-00019.safetensors",
171
+ "model.layers.13.block_sparse_moe.experts.2.w3.weight": "model-00008-of-00019.safetensors",
172
+ "model.layers.13.block_sparse_moe.experts.3.w1.weight": "model-00008-of-00019.safetensors",
173
+ "model.layers.13.block_sparse_moe.experts.3.w2.weight": "model-00008-of-00019.safetensors",
174
+ "model.layers.13.block_sparse_moe.experts.3.w3.weight": "model-00008-of-00019.safetensors",
175
+ "model.layers.13.block_sparse_moe.experts.4.w1.weight": "model-00008-of-00019.safetensors",
176
+ "model.layers.13.block_sparse_moe.experts.4.w2.weight": "model-00009-of-00019.safetensors",
177
+ "model.layers.13.block_sparse_moe.experts.4.w3.weight": "model-00009-of-00019.safetensors",
178
+ "model.layers.13.block_sparse_moe.experts.5.w1.weight": "model-00009-of-00019.safetensors",
179
+ "model.layers.13.block_sparse_moe.experts.5.w2.weight": "model-00009-of-00019.safetensors",
180
+ "model.layers.13.block_sparse_moe.experts.5.w3.weight": "model-00009-of-00019.safetensors",
181
+ "model.layers.13.block_sparse_moe.experts.6.w1.weight": "model-00009-of-00019.safetensors",
182
+ "model.layers.13.block_sparse_moe.experts.6.w2.weight": "model-00009-of-00019.safetensors",
183
+ "model.layers.13.block_sparse_moe.experts.6.w3.weight": "model-00009-of-00019.safetensors",
184
+ "model.layers.13.block_sparse_moe.experts.7.w1.weight": "model-00009-of-00019.safetensors",
185
+ "model.layers.13.block_sparse_moe.experts.7.w2.weight": "model-00009-of-00019.safetensors",
186
+ "model.layers.13.block_sparse_moe.experts.7.w3.weight": "model-00009-of-00019.safetensors",
187
+ "model.layers.13.block_sparse_moe.gate.weight": "model-00008-of-00019.safetensors",
188
+ "model.layers.13.input_layernorm.weight": "model-00009-of-00019.safetensors",
189
+ "model.layers.13.post_attention_layernorm.weight": "model-00009-of-00019.safetensors",
190
+ "model.layers.13.self_attn.k_proj.weight": "model-00008-of-00019.safetensors",
191
+ "model.layers.13.self_attn.o_proj.weight": "model-00008-of-00019.safetensors",
192
+ "model.layers.13.self_attn.q_proj.weight": "model-00008-of-00019.safetensors",
193
+ "model.layers.13.self_attn.v_proj.weight": "model-00008-of-00019.safetensors",
194
+ "model.layers.14.block_sparse_moe.experts.0.w1.weight": "model-00009-of-00019.safetensors",
195
+ "model.layers.14.block_sparse_moe.experts.0.w2.weight": "model-00009-of-00019.safetensors",
196
+ "model.layers.14.block_sparse_moe.experts.0.w3.weight": "model-00009-of-00019.safetensors",
197
+ "model.layers.14.block_sparse_moe.experts.1.w1.weight": "model-00009-of-00019.safetensors",
198
+ "model.layers.14.block_sparse_moe.experts.1.w2.weight": "model-00009-of-00019.safetensors",
199
+ "model.layers.14.block_sparse_moe.experts.1.w3.weight": "model-00009-of-00019.safetensors",
200
+ "model.layers.14.block_sparse_moe.experts.2.w1.weight": "model-00009-of-00019.safetensors",
201
+ "model.layers.14.block_sparse_moe.experts.2.w2.weight": "model-00009-of-00019.safetensors",
202
+ "model.layers.14.block_sparse_moe.experts.2.w3.weight": "model-00009-of-00019.safetensors",
203
+ "model.layers.14.block_sparse_moe.experts.3.w1.weight": "model-00009-of-00019.safetensors",
204
+ "model.layers.14.block_sparse_moe.experts.3.w2.weight": "model-00009-of-00019.safetensors",
205
+ "model.layers.14.block_sparse_moe.experts.3.w3.weight": "model-00009-of-00019.safetensors",
206
+ "model.layers.14.block_sparse_moe.experts.4.w1.weight": "model-00009-of-00019.safetensors",
207
+ "model.layers.14.block_sparse_moe.experts.4.w2.weight": "model-00009-of-00019.safetensors",
208
+ "model.layers.14.block_sparse_moe.experts.4.w3.weight": "model-00009-of-00019.safetensors",
209
+ "model.layers.14.block_sparse_moe.experts.5.w1.weight": "model-00009-of-00019.safetensors",
210
+ "model.layers.14.block_sparse_moe.experts.5.w2.weight": "model-00009-of-00019.safetensors",
211
+ "model.layers.14.block_sparse_moe.experts.5.w3.weight": "model-00009-of-00019.safetensors",
212
+ "model.layers.14.block_sparse_moe.experts.6.w1.weight": "model-00009-of-00019.safetensors",
213
+ "model.layers.14.block_sparse_moe.experts.6.w2.weight": "model-00009-of-00019.safetensors",
214
+ "model.layers.14.block_sparse_moe.experts.6.w3.weight": "model-00009-of-00019.safetensors",
215
+ "model.layers.14.block_sparse_moe.experts.7.w1.weight": "model-00009-of-00019.safetensors",
216
+ "model.layers.14.block_sparse_moe.experts.7.w2.weight": "model-00009-of-00019.safetensors",
217
+ "model.layers.14.block_sparse_moe.experts.7.w3.weight": "model-00009-of-00019.safetensors",
218
+ "model.layers.14.block_sparse_moe.gate.weight": "model-00009-of-00019.safetensors",
219
+ "model.layers.14.input_layernorm.weight": "model-00009-of-00019.safetensors",
220
+ "model.layers.14.post_attention_layernorm.weight": "model-00009-of-00019.safetensors",
221
+ "model.layers.14.self_attn.k_proj.weight": "model-00009-of-00019.safetensors",
222
+ "model.layers.14.self_attn.o_proj.weight": "model-00009-of-00019.safetensors",
223
+ "model.layers.14.self_attn.q_proj.weight": "model-00009-of-00019.safetensors",
224
+ "model.layers.14.self_attn.v_proj.weight": "model-00009-of-00019.safetensors",
225
+ "model.layers.15.block_sparse_moe.experts.0.w1.weight": "model-00009-of-00019.safetensors",
226
+ "model.layers.15.block_sparse_moe.experts.0.w2.weight": "model-00009-of-00019.safetensors",
227
+ "model.layers.15.block_sparse_moe.experts.0.w3.weight": "model-00009-of-00019.safetensors",
228
+ "model.layers.15.block_sparse_moe.experts.1.w1.weight": "model-00009-of-00019.safetensors",
229
+ "model.layers.15.block_sparse_moe.experts.1.w2.weight": "model-00009-of-00019.safetensors",
230
+ "model.layers.15.block_sparse_moe.experts.1.w3.weight": "model-00009-of-00019.safetensors",
231
+ "model.layers.15.block_sparse_moe.experts.2.w1.weight": "model-00010-of-00019.safetensors",
232
+ "model.layers.15.block_sparse_moe.experts.2.w2.weight": "model-00010-of-00019.safetensors",
233
+ "model.layers.15.block_sparse_moe.experts.2.w3.weight": "model-00010-of-00019.safetensors",
234
+ "model.layers.15.block_sparse_moe.experts.3.w1.weight": "model-00010-of-00019.safetensors",
235
+ "model.layers.15.block_sparse_moe.experts.3.w2.weight": "model-00010-of-00019.safetensors",
236
+ "model.layers.15.block_sparse_moe.experts.3.w3.weight": "model-00010-of-00019.safetensors",
237
+ "model.layers.15.block_sparse_moe.experts.4.w1.weight": "model-00010-of-00019.safetensors",
238
+ "model.layers.15.block_sparse_moe.experts.4.w2.weight": "model-00010-of-00019.safetensors",
239
+ "model.layers.15.block_sparse_moe.experts.4.w3.weight": "model-00010-of-00019.safetensors",
240
+ "model.layers.15.block_sparse_moe.experts.5.w1.weight": "model-00010-of-00019.safetensors",
241
+ "model.layers.15.block_sparse_moe.experts.5.w2.weight": "model-00010-of-00019.safetensors",
242
+ "model.layers.15.block_sparse_moe.experts.5.w3.weight": "model-00010-of-00019.safetensors",
243
+ "model.layers.15.block_sparse_moe.experts.6.w1.weight": "model-00010-of-00019.safetensors",
244
+ "model.layers.15.block_sparse_moe.experts.6.w2.weight": "model-00010-of-00019.safetensors",
245
+ "model.layers.15.block_sparse_moe.experts.6.w3.weight": "model-00010-of-00019.safetensors",
246
+ "model.layers.15.block_sparse_moe.experts.7.w1.weight": "model-00010-of-00019.safetensors",
247
+ "model.layers.15.block_sparse_moe.experts.7.w2.weight": "model-00010-of-00019.safetensors",
248
+ "model.layers.15.block_sparse_moe.experts.7.w3.weight": "model-00010-of-00019.safetensors",
249
+ "model.layers.15.block_sparse_moe.gate.weight": "model-00009-of-00019.safetensors",
250
+ "model.layers.15.input_layernorm.weight": "model-00010-of-00019.safetensors",
251
+ "model.layers.15.post_attention_layernorm.weight": "model-00010-of-00019.safetensors",
252
+ "model.layers.15.self_attn.k_proj.weight": "model-00009-of-00019.safetensors",
253
+ "model.layers.15.self_attn.o_proj.weight": "model-00009-of-00019.safetensors",
254
+ "model.layers.15.self_attn.q_proj.weight": "model-00009-of-00019.safetensors",
255
+ "model.layers.15.self_attn.v_proj.weight": "model-00009-of-00019.safetensors",
256
+ "model.layers.16.block_sparse_moe.experts.0.w1.weight": "model-00010-of-00019.safetensors",
257
+ "model.layers.16.block_sparse_moe.experts.0.w2.weight": "model-00010-of-00019.safetensors",
258
+ "model.layers.16.block_sparse_moe.experts.0.w3.weight": "model-00010-of-00019.safetensors",
259
+ "model.layers.16.block_sparse_moe.experts.1.w1.weight": "model-00010-of-00019.safetensors",
260
+ "model.layers.16.block_sparse_moe.experts.1.w2.weight": "model-00010-of-00019.safetensors",
261
+ "model.layers.16.block_sparse_moe.experts.1.w3.weight": "model-00010-of-00019.safetensors",
262
+ "model.layers.16.block_sparse_moe.experts.2.w1.weight": "model-00010-of-00019.safetensors",
263
+ "model.layers.16.block_sparse_moe.experts.2.w2.weight": "model-00010-of-00019.safetensors",
264
+ "model.layers.16.block_sparse_moe.experts.2.w3.weight": "model-00010-of-00019.safetensors",
265
+ "model.layers.16.block_sparse_moe.experts.3.w1.weight": "model-00010-of-00019.safetensors",
266
+ "model.layers.16.block_sparse_moe.experts.3.w2.weight": "model-00010-of-00019.safetensors",
267
+ "model.layers.16.block_sparse_moe.experts.3.w3.weight": "model-00010-of-00019.safetensors",
268
+ "model.layers.16.block_sparse_moe.experts.4.w1.weight": "model-00010-of-00019.safetensors",
269
+ "model.layers.16.block_sparse_moe.experts.4.w2.weight": "model-00010-of-00019.safetensors",
270
+ "model.layers.16.block_sparse_moe.experts.4.w3.weight": "model-00010-of-00019.safetensors",
271
+ "model.layers.16.block_sparse_moe.experts.5.w1.weight": "model-00010-of-00019.safetensors",
272
+ "model.layers.16.block_sparse_moe.experts.5.w2.weight": "model-00010-of-00019.safetensors",
273
+ "model.layers.16.block_sparse_moe.experts.5.w3.weight": "model-00010-of-00019.safetensors",
274
+ "model.layers.16.block_sparse_moe.experts.6.w1.weight": "model-00010-of-00019.safetensors",
275
+ "model.layers.16.block_sparse_moe.experts.6.w2.weight": "model-00010-of-00019.safetensors",
276
+ "model.layers.16.block_sparse_moe.experts.6.w3.weight": "model-00010-of-00019.safetensors",
277
+ "model.layers.16.block_sparse_moe.experts.7.w1.weight": "model-00010-of-00019.safetensors",
278
+ "model.layers.16.block_sparse_moe.experts.7.w2.weight": "model-00010-of-00019.safetensors",
279
+ "model.layers.16.block_sparse_moe.experts.7.w3.weight": "model-00011-of-00019.safetensors",
280
+ "model.layers.16.block_sparse_moe.gate.weight": "model-00010-of-00019.safetensors",
281
+ "model.layers.16.input_layernorm.weight": "model-00011-of-00019.safetensors",
282
+ "model.layers.16.post_attention_layernorm.weight": "model-00011-of-00019.safetensors",
283
+ "model.layers.16.self_attn.k_proj.weight": "model-00010-of-00019.safetensors",
284
+ "model.layers.16.self_attn.o_proj.weight": "model-00010-of-00019.safetensors",
285
+ "model.layers.16.self_attn.q_proj.weight": "model-00010-of-00019.safetensors",
286
+ "model.layers.16.self_attn.v_proj.weight": "model-00010-of-00019.safetensors",
287
+ "model.layers.17.block_sparse_moe.experts.0.w1.weight": "model-00011-of-00019.safetensors",
288
+ "model.layers.17.block_sparse_moe.experts.0.w2.weight": "model-00011-of-00019.safetensors",
289
+ "model.layers.17.block_sparse_moe.experts.0.w3.weight": "model-00011-of-00019.safetensors",
290
+ "model.layers.17.block_sparse_moe.experts.1.w1.weight": "model-00011-of-00019.safetensors",
291
+ "model.layers.17.block_sparse_moe.experts.1.w2.weight": "model-00011-of-00019.safetensors",
292
+ "model.layers.17.block_sparse_moe.experts.1.w3.weight": "model-00011-of-00019.safetensors",
293
+ "model.layers.17.block_sparse_moe.experts.2.w1.weight": "model-00011-of-00019.safetensors",
294
+ "model.layers.17.block_sparse_moe.experts.2.w2.weight": "model-00011-of-00019.safetensors",
295
+ "model.layers.17.block_sparse_moe.experts.2.w3.weight": "model-00011-of-00019.safetensors",
296
+ "model.layers.17.block_sparse_moe.experts.3.w1.weight": "model-00011-of-00019.safetensors",
297
+ "model.layers.17.block_sparse_moe.experts.3.w2.weight": "model-00011-of-00019.safetensors",
298
+ "model.layers.17.block_sparse_moe.experts.3.w3.weight": "model-00011-of-00019.safetensors",
299
+ "model.layers.17.block_sparse_moe.experts.4.w1.weight": "model-00011-of-00019.safetensors",
300
+ "model.layers.17.block_sparse_moe.experts.4.w2.weight": "model-00011-of-00019.safetensors",
301
+ "model.layers.17.block_sparse_moe.experts.4.w3.weight": "model-00011-of-00019.safetensors",
302
+ "model.layers.17.block_sparse_moe.experts.5.w1.weight": "model-00011-of-00019.safetensors",
303
+ "model.layers.17.block_sparse_moe.experts.5.w2.weight": "model-00011-of-00019.safetensors",
304
+ "model.layers.17.block_sparse_moe.experts.5.w3.weight": "model-00011-of-00019.safetensors",
305
+ "model.layers.17.block_sparse_moe.experts.6.w1.weight": "model-00011-of-00019.safetensors",
306
+ "model.layers.17.block_sparse_moe.experts.6.w2.weight": "model-00011-of-00019.safetensors",
307
+ "model.layers.17.block_sparse_moe.experts.6.w3.weight": "model-00011-of-00019.safetensors",
308
+ "model.layers.17.block_sparse_moe.experts.7.w1.weight": "model-00011-of-00019.safetensors",
309
+ "model.layers.17.block_sparse_moe.experts.7.w2.weight": "model-00011-of-00019.safetensors",
310
+ "model.layers.17.block_sparse_moe.experts.7.w3.weight": "model-00011-of-00019.safetensors",
311
+ "model.layers.17.block_sparse_moe.gate.weight": "model-00011-of-00019.safetensors",
312
+ "model.layers.17.input_layernorm.weight": "model-00011-of-00019.safetensors",
313
+ "model.layers.17.post_attention_layernorm.weight": "model-00011-of-00019.safetensors",
314
+ "model.layers.17.self_attn.k_proj.weight": "model-00011-of-00019.safetensors",
315
+ "model.layers.17.self_attn.o_proj.weight": "model-00011-of-00019.safetensors",
316
+ "model.layers.17.self_attn.q_proj.weight": "model-00011-of-00019.safetensors",
317
+ "model.layers.17.self_attn.v_proj.weight": "model-00011-of-00019.safetensors",
318
+ "model.layers.18.block_sparse_moe.experts.0.w1.weight": "model-00011-of-00019.safetensors",
319
+ "model.layers.18.block_sparse_moe.experts.0.w2.weight": "model-00011-of-00019.safetensors",
320
+ "model.layers.18.block_sparse_moe.experts.0.w3.weight": "model-00011-of-00019.safetensors",
321
+ "model.layers.18.block_sparse_moe.experts.1.w1.weight": "model-00011-of-00019.safetensors",
322
+ "model.layers.18.block_sparse_moe.experts.1.w2.weight": "model-00011-of-00019.safetensors",
323
+ "model.layers.18.block_sparse_moe.experts.1.w3.weight": "model-00011-of-00019.safetensors",
324
+ "model.layers.18.block_sparse_moe.experts.2.w1.weight": "model-00011-of-00019.safetensors",
325
+ "model.layers.18.block_sparse_moe.experts.2.w2.weight": "model-00011-of-00019.safetensors",
326
+ "model.layers.18.block_sparse_moe.experts.2.w3.weight": "model-00011-of-00019.safetensors",
327
+ "model.layers.18.block_sparse_moe.experts.3.w1.weight": "model-00011-of-00019.safetensors",
328
+ "model.layers.18.block_sparse_moe.experts.3.w2.weight": "model-00011-of-00019.safetensors",
329
+ "model.layers.18.block_sparse_moe.experts.3.w3.weight": "model-00011-of-00019.safetensors",
330
+ "model.layers.18.block_sparse_moe.experts.4.w1.weight": "model-00011-of-00019.safetensors",
331
+ "model.layers.18.block_sparse_moe.experts.4.w2.weight": "model-00011-of-00019.safetensors",
332
+ "model.layers.18.block_sparse_moe.experts.4.w3.weight": "model-00011-of-00019.safetensors",
333
+ "model.layers.18.block_sparse_moe.experts.5.w1.weight": "model-00011-of-00019.safetensors",
334
+ "model.layers.18.block_sparse_moe.experts.5.w2.weight": "model-00012-of-00019.safetensors",
335
+ "model.layers.18.block_sparse_moe.experts.5.w3.weight": "model-00012-of-00019.safetensors",
336
+ "model.layers.18.block_sparse_moe.experts.6.w1.weight": "model-00012-of-00019.safetensors",
337
+ "model.layers.18.block_sparse_moe.experts.6.w2.weight": "model-00012-of-00019.safetensors",
338
+ "model.layers.18.block_sparse_moe.experts.6.w3.weight": "model-00012-of-00019.safetensors",
339
+ "model.layers.18.block_sparse_moe.experts.7.w1.weight": "model-00012-of-00019.safetensors",
340
+ "model.layers.18.block_sparse_moe.experts.7.w2.weight": "model-00012-of-00019.safetensors",
341
+ "model.layers.18.block_sparse_moe.experts.7.w3.weight": "model-00012-of-00019.safetensors",
342
+ "model.layers.18.block_sparse_moe.gate.weight": "model-00011-of-00019.safetensors",
343
+ "model.layers.18.input_layernorm.weight": "model-00012-of-00019.safetensors",
344
+ "model.layers.18.post_attention_layernorm.weight": "model-00012-of-00019.safetensors",
345
+ "model.layers.18.self_attn.k_proj.weight": "model-00011-of-00019.safetensors",
346
+ "model.layers.18.self_attn.o_proj.weight": "model-00011-of-00019.safetensors",
347
+ "model.layers.18.self_attn.q_proj.weight": "model-00011-of-00019.safetensors",
348
+ "model.layers.18.self_attn.v_proj.weight": "model-00011-of-00019.safetensors",
349
+ "model.layers.19.block_sparse_moe.experts.0.w1.weight": "model-00012-of-00019.safetensors",
350
+ "model.layers.19.block_sparse_moe.experts.0.w2.weight": "model-00012-of-00019.safetensors",
351
+ "model.layers.19.block_sparse_moe.experts.0.w3.weight": "model-00012-of-00019.safetensors",
352
+ "model.layers.19.block_sparse_moe.experts.1.w1.weight": "model-00012-of-00019.safetensors",
353
+ "model.layers.19.block_sparse_moe.experts.1.w2.weight": "model-00012-of-00019.safetensors",
354
+ "model.layers.19.block_sparse_moe.experts.1.w3.weight": "model-00012-of-00019.safetensors",
355
+ "model.layers.19.block_sparse_moe.experts.2.w1.weight": "model-00012-of-00019.safetensors",
356
+ "model.layers.19.block_sparse_moe.experts.2.w2.weight": "model-00012-of-00019.safetensors",
357
+ "model.layers.19.block_sparse_moe.experts.2.w3.weight": "model-00012-of-00019.safetensors",
358
+ "model.layers.19.block_sparse_moe.experts.3.w1.weight": "model-00012-of-00019.safetensors",
359
+ "model.layers.19.block_sparse_moe.experts.3.w2.weight": "model-00012-of-00019.safetensors",
360
+ "model.layers.19.block_sparse_moe.experts.3.w3.weight": "model-00012-of-00019.safetensors",
361
+ "model.layers.19.block_sparse_moe.experts.4.w1.weight": "model-00012-of-00019.safetensors",
362
+ "model.layers.19.block_sparse_moe.experts.4.w2.weight": "model-00012-of-00019.safetensors",
363
+ "model.layers.19.block_sparse_moe.experts.4.w3.weight": "model-00012-of-00019.safetensors",
364
+ "model.layers.19.block_sparse_moe.experts.5.w1.weight": "model-00012-of-00019.safetensors",
365
+ "model.layers.19.block_sparse_moe.experts.5.w2.weight": "model-00012-of-00019.safetensors",
366
+ "model.layers.19.block_sparse_moe.experts.5.w3.weight": "model-00012-of-00019.safetensors",
367
+ "model.layers.19.block_sparse_moe.experts.6.w1.weight": "model-00012-of-00019.safetensors",
368
+ "model.layers.19.block_sparse_moe.experts.6.w2.weight": "model-00012-of-00019.safetensors",
369
+ "model.layers.19.block_sparse_moe.experts.6.w3.weight": "model-00012-of-00019.safetensors",
370
+ "model.layers.19.block_sparse_moe.experts.7.w1.weight": "model-00012-of-00019.safetensors",
371
+ "model.layers.19.block_sparse_moe.experts.7.w2.weight": "model-00012-of-00019.safetensors",
372
+ "model.layers.19.block_sparse_moe.experts.7.w3.weight": "model-00012-of-00019.safetensors",
373
+ "model.layers.19.block_sparse_moe.gate.weight": "model-00012-of-00019.safetensors",
374
+ "model.layers.19.input_layernorm.weight": "model-00012-of-00019.safetensors",
375
+ "model.layers.19.post_attention_layernorm.weight": "model-00012-of-00019.safetensors",
376
+ "model.layers.19.self_attn.k_proj.weight": "model-00012-of-00019.safetensors",
377
+ "model.layers.19.self_attn.o_proj.weight": "model-00012-of-00019.safetensors",
378
+ "model.layers.19.self_attn.q_proj.weight": "model-00012-of-00019.safetensors",
379
+ "model.layers.19.self_attn.v_proj.weight": "model-00012-of-00019.safetensors",
380
+ "model.layers.2.block_sparse_moe.experts.0.w1.weight": "model-00002-of-00019.safetensors",
381
+ "model.layers.2.block_sparse_moe.experts.0.w2.weight": "model-00002-of-00019.safetensors",
382
+ "model.layers.2.block_sparse_moe.experts.0.w3.weight": "model-00002-of-00019.safetensors",
383
+ "model.layers.2.block_sparse_moe.experts.1.w1.weight": "model-00002-of-00019.safetensors",
384
+ "model.layers.2.block_sparse_moe.experts.1.w2.weight": "model-00002-of-00019.safetensors",
385
+ "model.layers.2.block_sparse_moe.experts.1.w3.weight": "model-00002-of-00019.safetensors",
386
+ "model.layers.2.block_sparse_moe.experts.2.w1.weight": "model-00002-of-00019.safetensors",
387
+ "model.layers.2.block_sparse_moe.experts.2.w2.weight": "model-00002-of-00019.safetensors",
388
+ "model.layers.2.block_sparse_moe.experts.2.w3.weight": "model-00002-of-00019.safetensors",
389
+ "model.layers.2.block_sparse_moe.experts.3.w1.weight": "model-00002-of-00019.safetensors",
390
+ "model.layers.2.block_sparse_moe.experts.3.w2.weight": "model-00002-of-00019.safetensors",
391
+ "model.layers.2.block_sparse_moe.experts.3.w3.weight": "model-00002-of-00019.safetensors",
392
+ "model.layers.2.block_sparse_moe.experts.4.w1.weight": "model-00002-of-00019.safetensors",
393
+ "model.layers.2.block_sparse_moe.experts.4.w2.weight": "model-00002-of-00019.safetensors",
394
+ "model.layers.2.block_sparse_moe.experts.4.w3.weight": "model-00002-of-00019.safetensors",
395
+ "model.layers.2.block_sparse_moe.experts.5.w1.weight": "model-00002-of-00019.safetensors",
396
+ "model.layers.2.block_sparse_moe.experts.5.w2.weight": "model-00002-of-00019.safetensors",
397
+ "model.layers.2.block_sparse_moe.experts.5.w3.weight": "model-00002-of-00019.safetensors",
398
+ "model.layers.2.block_sparse_moe.experts.6.w1.weight": "model-00002-of-00019.safetensors",
399
+ "model.layers.2.block_sparse_moe.experts.6.w2.weight": "model-00002-of-00019.safetensors",
400
+ "model.layers.2.block_sparse_moe.experts.6.w3.weight": "model-00002-of-00019.safetensors",
401
+ "model.layers.2.block_sparse_moe.experts.7.w1.weight": "model-00002-of-00019.safetensors",
402
+ "model.layers.2.block_sparse_moe.experts.7.w2.weight": "model-00002-of-00019.safetensors",
403
+ "model.layers.2.block_sparse_moe.experts.7.w3.weight": "model-00002-of-00019.safetensors",
404
+ "model.layers.2.block_sparse_moe.gate.weight": "model-00002-of-00019.safetensors",
405
+ "model.layers.2.input_layernorm.weight": "model-00002-of-00019.safetensors",
406
+ "model.layers.2.post_attention_layernorm.weight": "model-00002-of-00019.safetensors",
407
+ "model.layers.2.self_attn.k_proj.weight": "model-00002-of-00019.safetensors",
408
+ "model.layers.2.self_attn.o_proj.weight": "model-00002-of-00019.safetensors",
409
+ "model.layers.2.self_attn.q_proj.weight": "model-00002-of-00019.safetensors",
410
+ "model.layers.2.self_attn.v_proj.weight": "model-00002-of-00019.safetensors",
411
+ "model.layers.20.block_sparse_moe.experts.0.w1.weight": "model-00012-of-00019.safetensors",
412
+ "model.layers.20.block_sparse_moe.experts.0.w2.weight": "model-00012-of-00019.safetensors",
413
+ "model.layers.20.block_sparse_moe.experts.0.w3.weight": "model-00012-of-00019.safetensors",
414
+ "model.layers.20.block_sparse_moe.experts.1.w1.weight": "model-00012-of-00019.safetensors",
415
+ "model.layers.20.block_sparse_moe.experts.1.w2.weight": "model-00012-of-00019.safetensors",
416
+ "model.layers.20.block_sparse_moe.experts.1.w3.weight": "model-00012-of-00019.safetensors",
417
+ "model.layers.20.block_sparse_moe.experts.2.w1.weight": "model-00012-of-00019.safetensors",
418
+ "model.layers.20.block_sparse_moe.experts.2.w2.weight": "model-00012-of-00019.safetensors",
419
+ "model.layers.20.block_sparse_moe.experts.2.w3.weight": "model-00012-of-00019.safetensors",
420
+ "model.layers.20.block_sparse_moe.experts.3.w1.weight": "model-00013-of-00019.safetensors",
421
+ "model.layers.20.block_sparse_moe.experts.3.w2.weight": "model-00013-of-00019.safetensors",
422
+ "model.layers.20.block_sparse_moe.experts.3.w3.weight": "model-00013-of-00019.safetensors",
423
+ "model.layers.20.block_sparse_moe.experts.4.w1.weight": "model-00013-of-00019.safetensors",
424
+ "model.layers.20.block_sparse_moe.experts.4.w2.weight": "model-00013-of-00019.safetensors",
425
+ "model.layers.20.block_sparse_moe.experts.4.w3.weight": "model-00013-of-00019.safetensors",
426
+ "model.layers.20.block_sparse_moe.experts.5.w1.weight": "model-00013-of-00019.safetensors",
427
+ "model.layers.20.block_sparse_moe.experts.5.w2.weight": "model-00013-of-00019.safetensors",
428
+ "model.layers.20.block_sparse_moe.experts.5.w3.weight": "model-00013-of-00019.safetensors",
429
+ "model.layers.20.block_sparse_moe.experts.6.w1.weight": "model-00013-of-00019.safetensors",
430
+ "model.layers.20.block_sparse_moe.experts.6.w2.weight": "model-00013-of-00019.safetensors",
431
+ "model.layers.20.block_sparse_moe.experts.6.w3.weight": "model-00013-of-00019.safetensors",
432
+ "model.layers.20.block_sparse_moe.experts.7.w1.weight": "model-00013-of-00019.safetensors",
433
+ "model.layers.20.block_sparse_moe.experts.7.w2.weight": "model-00013-of-00019.safetensors",
434
+ "model.layers.20.block_sparse_moe.experts.7.w3.weight": "model-00013-of-00019.safetensors",
435
+ "model.layers.20.block_sparse_moe.gate.weight": "model-00012-of-00019.safetensors",
436
+ "model.layers.20.input_layernorm.weight": "model-00013-of-00019.safetensors",
437
+ "model.layers.20.post_attention_layernorm.weight": "model-00013-of-00019.safetensors",
438
+ "model.layers.20.self_attn.k_proj.weight": "model-00012-of-00019.safetensors",
439
+ "model.layers.20.self_attn.o_proj.weight": "model-00012-of-00019.safetensors",
440
+ "model.layers.20.self_attn.q_proj.weight": "model-00012-of-00019.safetensors",
441
+ "model.layers.20.self_attn.v_proj.weight": "model-00012-of-00019.safetensors",
442
+ "model.layers.21.block_sparse_moe.experts.0.w1.weight": "model-00013-of-00019.safetensors",
443
+ "model.layers.21.block_sparse_moe.experts.0.w2.weight": "model-00013-of-00019.safetensors",
444
+ "model.layers.21.block_sparse_moe.experts.0.w3.weight": "model-00013-of-00019.safetensors",
445
+ "model.layers.21.block_sparse_moe.experts.1.w1.weight": "model-00013-of-00019.safetensors",
446
+ "model.layers.21.block_sparse_moe.experts.1.w2.weight": "model-00013-of-00019.safetensors",
447
+ "model.layers.21.block_sparse_moe.experts.1.w3.weight": "model-00013-of-00019.safetensors",
448
+ "model.layers.21.block_sparse_moe.experts.2.w1.weight": "model-00013-of-00019.safetensors",
449
+ "model.layers.21.block_sparse_moe.experts.2.w2.weight": "model-00013-of-00019.safetensors",
450
+ "model.layers.21.block_sparse_moe.experts.2.w3.weight": "model-00013-of-00019.safetensors",
451
+ "model.layers.21.block_sparse_moe.experts.3.w1.weight": "model-00013-of-00019.safetensors",
452
+ "model.layers.21.block_sparse_moe.experts.3.w2.weight": "model-00013-of-00019.safetensors",
453
+ "model.layers.21.block_sparse_moe.experts.3.w3.weight": "model-00013-of-00019.safetensors",
454
+ "model.layers.21.block_sparse_moe.experts.4.w1.weight": "model-00013-of-00019.safetensors",
455
+ "model.layers.21.block_sparse_moe.experts.4.w2.weight": "model-00013-of-00019.safetensors",
456
+ "model.layers.21.block_sparse_moe.experts.4.w3.weight": "model-00013-of-00019.safetensors",
457
+ "model.layers.21.block_sparse_moe.experts.5.w1.weight": "model-00013-of-00019.safetensors",
458
+ "model.layers.21.block_sparse_moe.experts.5.w2.weight": "model-00013-of-00019.safetensors",
459
+ "model.layers.21.block_sparse_moe.experts.5.w3.weight": "model-00013-of-00019.safetensors",
460
+ "model.layers.21.block_sparse_moe.experts.6.w1.weight": "model-00013-of-00019.safetensors",
461
+ "model.layers.21.block_sparse_moe.experts.6.w2.weight": "model-00013-of-00019.safetensors",
462
+ "model.layers.21.block_sparse_moe.experts.6.w3.weight": "model-00013-of-00019.safetensors",
463
+ "model.layers.21.block_sparse_moe.experts.7.w1.weight": "model-00013-of-00019.safetensors",
464
+ "model.layers.21.block_sparse_moe.experts.7.w2.weight": "model-00013-of-00019.safetensors",
465
+ "model.layers.21.block_sparse_moe.experts.7.w3.weight": "model-00013-of-00019.safetensors",
466
+ "model.layers.21.block_sparse_moe.gate.weight": "model-00013-of-00019.safetensors",
467
+ "model.layers.21.input_layernorm.weight": "model-00013-of-00019.safetensors",
468
+ "model.layers.21.post_attention_layernorm.weight": "model-00013-of-00019.safetensors",
469
+ "model.layers.21.self_attn.k_proj.weight": "model-00013-of-00019.safetensors",
470
+ "model.layers.21.self_attn.o_proj.weight": "model-00013-of-00019.safetensors",
471
+ "model.layers.21.self_attn.q_proj.weight": "model-00013-of-00019.safetensors",
472
+ "model.layers.21.self_attn.v_proj.weight": "model-00013-of-00019.safetensors",
473
+ "model.layers.22.block_sparse_moe.experts.0.w1.weight": "model-00013-of-00019.safetensors",
474
+ "model.layers.22.block_sparse_moe.experts.0.w2.weight": "model-00013-of-00019.safetensors",
475
+ "model.layers.22.block_sparse_moe.experts.0.w3.weight": "model-00014-of-00019.safetensors",
476
+ "model.layers.22.block_sparse_moe.experts.1.w1.weight": "model-00014-of-00019.safetensors",
477
+ "model.layers.22.block_sparse_moe.experts.1.w2.weight": "model-00014-of-00019.safetensors",
478
+ "model.layers.22.block_sparse_moe.experts.1.w3.weight": "model-00014-of-00019.safetensors",
479
+ "model.layers.22.block_sparse_moe.experts.2.w1.weight": "model-00014-of-00019.safetensors",
480
+ "model.layers.22.block_sparse_moe.experts.2.w2.weight": "model-00014-of-00019.safetensors",
481
+ "model.layers.22.block_sparse_moe.experts.2.w3.weight": "model-00014-of-00019.safetensors",
482
+ "model.layers.22.block_sparse_moe.experts.3.w1.weight": "model-00014-of-00019.safetensors",
483
+ "model.layers.22.block_sparse_moe.experts.3.w2.weight": "model-00014-of-00019.safetensors",
484
+ "model.layers.22.block_sparse_moe.experts.3.w3.weight": "model-00014-of-00019.safetensors",
485
+ "model.layers.22.block_sparse_moe.experts.4.w1.weight": "model-00014-of-00019.safetensors",
486
+ "model.layers.22.block_sparse_moe.experts.4.w2.weight": "model-00014-of-00019.safetensors",
487
+ "model.layers.22.block_sparse_moe.experts.4.w3.weight": "model-00014-of-00019.safetensors",
488
+ "model.layers.22.block_sparse_moe.experts.5.w1.weight": "model-00014-of-00019.safetensors",
489
+ "model.layers.22.block_sparse_moe.experts.5.w2.weight": "model-00014-of-00019.safetensors",
490
+ "model.layers.22.block_sparse_moe.experts.5.w3.weight": "model-00014-of-00019.safetensors",
491
+ "model.layers.22.block_sparse_moe.experts.6.w1.weight": "model-00014-of-00019.safetensors",
492
+ "model.layers.22.block_sparse_moe.experts.6.w2.weight": "model-00014-of-00019.safetensors",
493
+ "model.layers.22.block_sparse_moe.experts.6.w3.weight": "model-00014-of-00019.safetensors",
494
+ "model.layers.22.block_sparse_moe.experts.7.w1.weight": "model-00014-of-00019.safetensors",
495
+ "model.layers.22.block_sparse_moe.experts.7.w2.weight": "model-00014-of-00019.safetensors",
496
+ "model.layers.22.block_sparse_moe.experts.7.w3.weight": "model-00014-of-00019.safetensors",
497
+ "model.layers.22.block_sparse_moe.gate.weight": "model-00013-of-00019.safetensors",
498
+ "model.layers.22.input_layernorm.weight": "model-00014-of-00019.safetensors",
499
+ "model.layers.22.post_attention_layernorm.weight": "model-00014-of-00019.safetensors",
500
+ "model.layers.22.self_attn.k_proj.weight": "model-00013-of-00019.safetensors",
501
+ "model.layers.22.self_attn.o_proj.weight": "model-00013-of-00019.safetensors",
502
+ "model.layers.22.self_attn.q_proj.weight": "model-00013-of-00019.safetensors",
503
+ "model.layers.22.self_attn.v_proj.weight": "model-00013-of-00019.safetensors",
504
+ "model.layers.23.block_sparse_moe.experts.0.w1.weight": "model-00014-of-00019.safetensors",
505
+ "model.layers.23.block_sparse_moe.experts.0.w2.weight": "model-00014-of-00019.safetensors",
506
+ "model.layers.23.block_sparse_moe.experts.0.w3.weight": "model-00014-of-00019.safetensors",
507
+ "model.layers.23.block_sparse_moe.experts.1.w1.weight": "model-00014-of-00019.safetensors",
508
+ "model.layers.23.block_sparse_moe.experts.1.w2.weight": "model-00014-of-00019.safetensors",
509
+ "model.layers.23.block_sparse_moe.experts.1.w3.weight": "model-00014-of-00019.safetensors",
510
+ "model.layers.23.block_sparse_moe.experts.2.w1.weight": "model-00014-of-00019.safetensors",
511
+ "model.layers.23.block_sparse_moe.experts.2.w2.weight": "model-00014-of-00019.safetensors",
512
+ "model.layers.23.block_sparse_moe.experts.2.w3.weight": "model-00014-of-00019.safetensors",
513
+ "model.layers.23.block_sparse_moe.experts.3.w1.weight": "model-00014-of-00019.safetensors",
514
+ "model.layers.23.block_sparse_moe.experts.3.w2.weight": "model-00014-of-00019.safetensors",
515
+ "model.layers.23.block_sparse_moe.experts.3.w3.weight": "model-00014-of-00019.safetensors",
516
+ "model.layers.23.block_sparse_moe.experts.4.w1.weight": "model-00014-of-00019.safetensors",
517
+ "model.layers.23.block_sparse_moe.experts.4.w2.weight": "model-00014-of-00019.safetensors",
518
+ "model.layers.23.block_sparse_moe.experts.4.w3.weight": "model-00014-of-00019.safetensors",
519
+ "model.layers.23.block_sparse_moe.experts.5.w1.weight": "model-00014-of-00019.safetensors",
520
+ "model.layers.23.block_sparse_moe.experts.5.w2.weight": "model-00014-of-00019.safetensors",
521
+ "model.layers.23.block_sparse_moe.experts.5.w3.weight": "model-00014-of-00019.safetensors",
522
+ "model.layers.23.block_sparse_moe.experts.6.w1.weight": "model-00014-of-00019.safetensors",
523
+ "model.layers.23.block_sparse_moe.experts.6.w2.weight": "model-00015-of-00019.safetensors",
524
+ "model.layers.23.block_sparse_moe.experts.6.w3.weight": "model-00015-of-00019.safetensors",
525
+ "model.layers.23.block_sparse_moe.experts.7.w1.weight": "model-00015-of-00019.safetensors",
526
+ "model.layers.23.block_sparse_moe.experts.7.w2.weight": "model-00015-of-00019.safetensors",
527
+ "model.layers.23.block_sparse_moe.experts.7.w3.weight": "model-00015-of-00019.safetensors",
528
+ "model.layers.23.block_sparse_moe.gate.weight": "model-00014-of-00019.safetensors",
529
+ "model.layers.23.input_layernorm.weight": "model-00015-of-00019.safetensors",
530
+ "model.layers.23.post_attention_layernorm.weight": "model-00015-of-00019.safetensors",
531
+ "model.layers.23.self_attn.k_proj.weight": "model-00014-of-00019.safetensors",
532
+ "model.layers.23.self_attn.o_proj.weight": "model-00014-of-00019.safetensors",
533
+ "model.layers.23.self_attn.q_proj.weight": "model-00014-of-00019.safetensors",
534
+ "model.layers.23.self_attn.v_proj.weight": "model-00014-of-00019.safetensors",
535
+ "model.layers.24.block_sparse_moe.experts.0.w1.weight": "model-00015-of-00019.safetensors",
536
+ "model.layers.24.block_sparse_moe.experts.0.w2.weight": "model-00015-of-00019.safetensors",
537
+ "model.layers.24.block_sparse_moe.experts.0.w3.weight": "model-00015-of-00019.safetensors",
538
+ "model.layers.24.block_sparse_moe.experts.1.w1.weight": "model-00015-of-00019.safetensors",
539
+ "model.layers.24.block_sparse_moe.experts.1.w2.weight": "model-00015-of-00019.safetensors",
540
+ "model.layers.24.block_sparse_moe.experts.1.w3.weight": "model-00015-of-00019.safetensors",
541
+ "model.layers.24.block_sparse_moe.experts.2.w1.weight": "model-00015-of-00019.safetensors",
542
+ "model.layers.24.block_sparse_moe.experts.2.w2.weight": "model-00015-of-00019.safetensors",
543
+ "model.layers.24.block_sparse_moe.experts.2.w3.weight": "model-00015-of-00019.safetensors",
544
+ "model.layers.24.block_sparse_moe.experts.3.w1.weight": "model-00015-of-00019.safetensors",
545
+ "model.layers.24.block_sparse_moe.experts.3.w2.weight": "model-00015-of-00019.safetensors",
546
+ "model.layers.24.block_sparse_moe.experts.3.w3.weight": "model-00015-of-00019.safetensors",
547
+ "model.layers.24.block_sparse_moe.experts.4.w1.weight": "model-00015-of-00019.safetensors",
548
+ "model.layers.24.block_sparse_moe.experts.4.w2.weight": "model-00015-of-00019.safetensors",
549
+ "model.layers.24.block_sparse_moe.experts.4.w3.weight": "model-00015-of-00019.safetensors",
550
+ "model.layers.24.block_sparse_moe.experts.5.w1.weight": "model-00015-of-00019.safetensors",
551
+ "model.layers.24.block_sparse_moe.experts.5.w2.weight": "model-00015-of-00019.safetensors",
552
+ "model.layers.24.block_sparse_moe.experts.5.w3.weight": "model-00015-of-00019.safetensors",
553
+ "model.layers.24.block_sparse_moe.experts.6.w1.weight": "model-00015-of-00019.safetensors",
554
+ "model.layers.24.block_sparse_moe.experts.6.w2.weight": "model-00015-of-00019.safetensors",
555
+ "model.layers.24.block_sparse_moe.experts.6.w3.weight": "model-00015-of-00019.safetensors",
556
+ "model.layers.24.block_sparse_moe.experts.7.w1.weight": "model-00015-of-00019.safetensors",
557
+ "model.layers.24.block_sparse_moe.experts.7.w2.weight": "model-00015-of-00019.safetensors",
558
+ "model.layers.24.block_sparse_moe.experts.7.w3.weight": "model-00015-of-00019.safetensors",
559
+ "model.layers.24.block_sparse_moe.gate.weight": "model-00015-of-00019.safetensors",
560
+ "model.layers.24.input_layernorm.weight": "model-00015-of-00019.safetensors",
561
+ "model.layers.24.post_attention_layernorm.weight": "model-00015-of-00019.safetensors",
562
+ "model.layers.24.self_attn.k_proj.weight": "model-00015-of-00019.safetensors",
563
+ "model.layers.24.self_attn.o_proj.weight": "model-00015-of-00019.safetensors",
564
+ "model.layers.24.self_attn.q_proj.weight": "model-00015-of-00019.safetensors",
565
+ "model.layers.24.self_attn.v_proj.weight": "model-00015-of-00019.safetensors",
566
+ "model.layers.25.block_sparse_moe.experts.0.w1.weight": "model-00015-of-00019.safetensors",
567
+ "model.layers.25.block_sparse_moe.experts.0.w2.weight": "model-00015-of-00019.safetensors",
568
+ "model.layers.25.block_sparse_moe.experts.0.w3.weight": "model-00015-of-00019.safetensors",
569
+ "model.layers.25.block_sparse_moe.experts.1.w1.weight": "model-00015-of-00019.safetensors",
570
+ "model.layers.25.block_sparse_moe.experts.1.w2.weight": "model-00015-of-00019.safetensors",
571
+ "model.layers.25.block_sparse_moe.experts.1.w3.weight": "model-00015-of-00019.safetensors",
572
+ "model.layers.25.block_sparse_moe.experts.2.w1.weight": "model-00015-of-00019.safetensors",
573
+ "model.layers.25.block_sparse_moe.experts.2.w2.weight": "model-00015-of-00019.safetensors",
574
+ "model.layers.25.block_sparse_moe.experts.2.w3.weight": "model-00015-of-00019.safetensors",
575
+ "model.layers.25.block_sparse_moe.experts.3.w1.weight": "model-00015-of-00019.safetensors",
576
+ "model.layers.25.block_sparse_moe.experts.3.w2.weight": "model-00015-of-00019.safetensors",
577
+ "model.layers.25.block_sparse_moe.experts.3.w3.weight": "model-00015-of-00019.safetensors",
578
+ "model.layers.25.block_sparse_moe.experts.4.w1.weight": "model-00016-of-00019.safetensors",
579
+ "model.layers.25.block_sparse_moe.experts.4.w2.weight": "model-00016-of-00019.safetensors",
580
+ "model.layers.25.block_sparse_moe.experts.4.w3.weight": "model-00016-of-00019.safetensors",
581
+ "model.layers.25.block_sparse_moe.experts.5.w1.weight": "model-00016-of-00019.safetensors",
582
+ "model.layers.25.block_sparse_moe.experts.5.w2.weight": "model-00016-of-00019.safetensors",
583
+ "model.layers.25.block_sparse_moe.experts.5.w3.weight": "model-00016-of-00019.safetensors",
584
+ "model.layers.25.block_sparse_moe.experts.6.w1.weight": "model-00016-of-00019.safetensors",
585
+ "model.layers.25.block_sparse_moe.experts.6.w2.weight": "model-00016-of-00019.safetensors",
586
+ "model.layers.25.block_sparse_moe.experts.6.w3.weight": "model-00016-of-00019.safetensors",
587
+ "model.layers.25.block_sparse_moe.experts.7.w1.weight": "model-00016-of-00019.safetensors",
588
+ "model.layers.25.block_sparse_moe.experts.7.w2.weight": "model-00016-of-00019.safetensors",
589
+ "model.layers.25.block_sparse_moe.experts.7.w3.weight": "model-00016-of-00019.safetensors",
590
+ "model.layers.25.block_sparse_moe.gate.weight": "model-00015-of-00019.safetensors",
591
+ "model.layers.25.input_layernorm.weight": "model-00016-of-00019.safetensors",
592
+ "model.layers.25.post_attention_layernorm.weight": "model-00016-of-00019.safetensors",
593
+ "model.layers.25.self_attn.k_proj.weight": "model-00015-of-00019.safetensors",
594
+ "model.layers.25.self_attn.o_proj.weight": "model-00015-of-00019.safetensors",
595
+ "model.layers.25.self_attn.q_proj.weight": "model-00015-of-00019.safetensors",
596
+ "model.layers.25.self_attn.v_proj.weight": "model-00015-of-00019.safetensors",
597
+ "model.layers.26.block_sparse_moe.experts.0.w1.weight": "model-00016-of-00019.safetensors",
598
+ "model.layers.26.block_sparse_moe.experts.0.w2.weight": "model-00016-of-00019.safetensors",
599
+ "model.layers.26.block_sparse_moe.experts.0.w3.weight": "model-00016-of-00019.safetensors",
600
+ "model.layers.26.block_sparse_moe.experts.1.w1.weight": "model-00016-of-00019.safetensors",
601
+ "model.layers.26.block_sparse_moe.experts.1.w2.weight": "model-00016-of-00019.safetensors",
602
+ "model.layers.26.block_sparse_moe.experts.1.w3.weight": "model-00016-of-00019.safetensors",
603
+ "model.layers.26.block_sparse_moe.experts.2.w1.weight": "model-00016-of-00019.safetensors",
604
+ "model.layers.26.block_sparse_moe.experts.2.w2.weight": "model-00016-of-00019.safetensors",
605
+ "model.layers.26.block_sparse_moe.experts.2.w3.weight": "model-00016-of-00019.safetensors",
606
+ "model.layers.26.block_sparse_moe.experts.3.w1.weight": "model-00016-of-00019.safetensors",
607
+ "model.layers.26.block_sparse_moe.experts.3.w2.weight": "model-00016-of-00019.safetensors",
608
+ "model.layers.26.block_sparse_moe.experts.3.w3.weight": "model-00016-of-00019.safetensors",
609
+ "model.layers.26.block_sparse_moe.experts.4.w1.weight": "model-00016-of-00019.safetensors",
610
+ "model.layers.26.block_sparse_moe.experts.4.w2.weight": "model-00016-of-00019.safetensors",
611
+ "model.layers.26.block_sparse_moe.experts.4.w3.weight": "model-00016-of-00019.safetensors",
612
+ "model.layers.26.block_sparse_moe.experts.5.w1.weight": "model-00016-of-00019.safetensors",
613
+ "model.layers.26.block_sparse_moe.experts.5.w2.weight": "model-00016-of-00019.safetensors",
614
+ "model.layers.26.block_sparse_moe.experts.5.w3.weight": "model-00016-of-00019.safetensors",
615
+ "model.layers.26.block_sparse_moe.experts.6.w1.weight": "model-00016-of-00019.safetensors",
616
+ "model.layers.26.block_sparse_moe.experts.6.w2.weight": "model-00016-of-00019.safetensors",
617
+ "model.layers.26.block_sparse_moe.experts.6.w3.weight": "model-00016-of-00019.safetensors",
618
+ "model.layers.26.block_sparse_moe.experts.7.w1.weight": "model-00016-of-00019.safetensors",
619
+ "model.layers.26.block_sparse_moe.experts.7.w2.weight": "model-00016-of-00019.safetensors",
620
+ "model.layers.26.block_sparse_moe.experts.7.w3.weight": "model-00016-of-00019.safetensors",
621
+ "model.layers.26.block_sparse_moe.gate.weight": "model-00016-of-00019.safetensors",
622
+ "model.layers.26.input_layernorm.weight": "model-00016-of-00019.safetensors",
623
+ "model.layers.26.post_attention_layernorm.weight": "model-00016-of-00019.safetensors",
624
+ "model.layers.26.self_attn.k_proj.weight": "model-00016-of-00019.safetensors",
625
+ "model.layers.26.self_attn.o_proj.weight": "model-00016-of-00019.safetensors",
626
+ "model.layers.26.self_attn.q_proj.weight": "model-00016-of-00019.safetensors",
627
+ "model.layers.26.self_attn.v_proj.weight": "model-00016-of-00019.safetensors",
628
+ "model.layers.27.block_sparse_moe.experts.0.w1.weight": "model-00016-of-00019.safetensors",
629
+ "model.layers.27.block_sparse_moe.experts.0.w2.weight": "model-00016-of-00019.safetensors",
630
+ "model.layers.27.block_sparse_moe.experts.0.w3.weight": "model-00016-of-00019.safetensors",
631
+ "model.layers.27.block_sparse_moe.experts.1.w1.weight": "model-00016-of-00019.safetensors",
632
+ "model.layers.27.block_sparse_moe.experts.1.w2.weight": "model-00016-of-00019.safetensors",
633
+ "model.layers.27.block_sparse_moe.experts.1.w3.weight": "model-00017-of-00019.safetensors",
634
+ "model.layers.27.block_sparse_moe.experts.2.w1.weight": "model-00017-of-00019.safetensors",
635
+ "model.layers.27.block_sparse_moe.experts.2.w2.weight": "model-00017-of-00019.safetensors",
636
+ "model.layers.27.block_sparse_moe.experts.2.w3.weight": "model-00017-of-00019.safetensors",
637
+ "model.layers.27.block_sparse_moe.experts.3.w1.weight": "model-00017-of-00019.safetensors",
638
+ "model.layers.27.block_sparse_moe.experts.3.w2.weight": "model-00017-of-00019.safetensors",
639
+ "model.layers.27.block_sparse_moe.experts.3.w3.weight": "model-00017-of-00019.safetensors",
640
+ "model.layers.27.block_sparse_moe.experts.4.w1.weight": "model-00017-of-00019.safetensors",
641
+ "model.layers.27.block_sparse_moe.experts.4.w2.weight": "model-00017-of-00019.safetensors",
642
+ "model.layers.27.block_sparse_moe.experts.4.w3.weight": "model-00017-of-00019.safetensors",
643
+ "model.layers.27.block_sparse_moe.experts.5.w1.weight": "model-00017-of-00019.safetensors",
644
+ "model.layers.27.block_sparse_moe.experts.5.w2.weight": "model-00017-of-00019.safetensors",
645
+ "model.layers.27.block_sparse_moe.experts.5.w3.weight": "model-00017-of-00019.safetensors",
646
+ "model.layers.27.block_sparse_moe.experts.6.w1.weight": "model-00017-of-00019.safetensors",
647
+ "model.layers.27.block_sparse_moe.experts.6.w2.weight": "model-00017-of-00019.safetensors",
648
+ "model.layers.27.block_sparse_moe.experts.6.w3.weight": "model-00017-of-00019.safetensors",
649
+ "model.layers.27.block_sparse_moe.experts.7.w1.weight": "model-00017-of-00019.safetensors",
650
+ "model.layers.27.block_sparse_moe.experts.7.w2.weight": "model-00017-of-00019.safetensors",
651
+ "model.layers.27.block_sparse_moe.experts.7.w3.weight": "model-00017-of-00019.safetensors",
652
+ "model.layers.27.block_sparse_moe.gate.weight": "model-00016-of-00019.safetensors",
653
+ "model.layers.27.input_layernorm.weight": "model-00017-of-00019.safetensors",
654
+ "model.layers.27.post_attention_layernorm.weight": "model-00017-of-00019.safetensors",
655
+ "model.layers.27.self_attn.k_proj.weight": "model-00016-of-00019.safetensors",
656
+ "model.layers.27.self_attn.o_proj.weight": "model-00016-of-00019.safetensors",
657
+ "model.layers.27.self_attn.q_proj.weight": "model-00016-of-00019.safetensors",
658
+ "model.layers.27.self_attn.v_proj.weight": "model-00016-of-00019.safetensors",
659
+ "model.layers.28.block_sparse_moe.experts.0.w1.weight": "model-00017-of-00019.safetensors",
660
+ "model.layers.28.block_sparse_moe.experts.0.w2.weight": "model-00017-of-00019.safetensors",
661
+ "model.layers.28.block_sparse_moe.experts.0.w3.weight": "model-00017-of-00019.safetensors",
662
+ "model.layers.28.block_sparse_moe.experts.1.w1.weight": "model-00017-of-00019.safetensors",
663
+ "model.layers.28.block_sparse_moe.experts.1.w2.weight": "model-00017-of-00019.safetensors",
664
+ "model.layers.28.block_sparse_moe.experts.1.w3.weight": "model-00017-of-00019.safetensors",
665
+ "model.layers.28.block_sparse_moe.experts.2.w1.weight": "model-00017-of-00019.safetensors",
666
+ "model.layers.28.block_sparse_moe.experts.2.w2.weight": "model-00017-of-00019.safetensors",
667
+ "model.layers.28.block_sparse_moe.experts.2.w3.weight": "model-00017-of-00019.safetensors",
668
+ "model.layers.28.block_sparse_moe.experts.3.w1.weight": "model-00017-of-00019.safetensors",
669
+ "model.layers.28.block_sparse_moe.experts.3.w2.weight": "model-00017-of-00019.safetensors",
670
+ "model.layers.28.block_sparse_moe.experts.3.w3.weight": "model-00017-of-00019.safetensors",
671
+ "model.layers.28.block_sparse_moe.experts.4.w1.weight": "model-00017-of-00019.safetensors",
672
+ "model.layers.28.block_sparse_moe.experts.4.w2.weight": "model-00017-of-00019.safetensors",
673
+ "model.layers.28.block_sparse_moe.experts.4.w3.weight": "model-00017-of-00019.safetensors",
674
+ "model.layers.28.block_sparse_moe.experts.5.w1.weight": "model-00017-of-00019.safetensors",
675
+ "model.layers.28.block_sparse_moe.experts.5.w2.weight": "model-00017-of-00019.safetensors",
676
+ "model.layers.28.block_sparse_moe.experts.5.w3.weight": "model-00017-of-00019.safetensors",
677
+ "model.layers.28.block_sparse_moe.experts.6.w1.weight": "model-00017-of-00019.safetensors",
678
+ "model.layers.28.block_sparse_moe.experts.6.w2.weight": "model-00017-of-00019.safetensors",
679
+ "model.layers.28.block_sparse_moe.experts.6.w3.weight": "model-00017-of-00019.safetensors",
680
+ "model.layers.28.block_sparse_moe.experts.7.w1.weight": "model-00017-of-00019.safetensors",
681
+ "model.layers.28.block_sparse_moe.experts.7.w2.weight": "model-00018-of-00019.safetensors",
682
+ "model.layers.28.block_sparse_moe.experts.7.w3.weight": "model-00018-of-00019.safetensors",
683
+ "model.layers.28.block_sparse_moe.gate.weight": "model-00017-of-00019.safetensors",
684
+ "model.layers.28.input_layernorm.weight": "model-00018-of-00019.safetensors",
685
+ "model.layers.28.post_attention_layernorm.weight": "model-00018-of-00019.safetensors",
686
+ "model.layers.28.self_attn.k_proj.weight": "model-00017-of-00019.safetensors",
687
+ "model.layers.28.self_attn.o_proj.weight": "model-00017-of-00019.safetensors",
688
+ "model.layers.28.self_attn.q_proj.weight": "model-00017-of-00019.safetensors",
689
+ "model.layers.28.self_attn.v_proj.weight": "model-00017-of-00019.safetensors",
690
+ "model.layers.29.block_sparse_moe.experts.0.w1.weight": "model-00018-of-00019.safetensors",
691
+ "model.layers.29.block_sparse_moe.experts.0.w2.weight": "model-00018-of-00019.safetensors",
692
+ "model.layers.29.block_sparse_moe.experts.0.w3.weight": "model-00018-of-00019.safetensors",
693
+ "model.layers.29.block_sparse_moe.experts.1.w1.weight": "model-00018-of-00019.safetensors",
694
+ "model.layers.29.block_sparse_moe.experts.1.w2.weight": "model-00018-of-00019.safetensors",
695
+ "model.layers.29.block_sparse_moe.experts.1.w3.weight": "model-00018-of-00019.safetensors",
696
+ "model.layers.29.block_sparse_moe.experts.2.w1.weight": "model-00018-of-00019.safetensors",
697
+ "model.layers.29.block_sparse_moe.experts.2.w2.weight": "model-00018-of-00019.safetensors",
698
+ "model.layers.29.block_sparse_moe.experts.2.w3.weight": "model-00018-of-00019.safetensors",
699
+ "model.layers.29.block_sparse_moe.experts.3.w1.weight": "model-00018-of-00019.safetensors",
700
+ "model.layers.29.block_sparse_moe.experts.3.w2.weight": "model-00018-of-00019.safetensors",
701
+ "model.layers.29.block_sparse_moe.experts.3.w3.weight": "model-00018-of-00019.safetensors",
702
+ "model.layers.29.block_sparse_moe.experts.4.w1.weight": "model-00018-of-00019.safetensors",
703
+ "model.layers.29.block_sparse_moe.experts.4.w2.weight": "model-00018-of-00019.safetensors",
704
+ "model.layers.29.block_sparse_moe.experts.4.w3.weight": "model-00018-of-00019.safetensors",
705
+ "model.layers.29.block_sparse_moe.experts.5.w1.weight": "model-00018-of-00019.safetensors",
706
+ "model.layers.29.block_sparse_moe.experts.5.w2.weight": "model-00018-of-00019.safetensors",
707
+ "model.layers.29.block_sparse_moe.experts.5.w3.weight": "model-00018-of-00019.safetensors",
708
+ "model.layers.29.block_sparse_moe.experts.6.w1.weight": "model-00018-of-00019.safetensors",
709
+ "model.layers.29.block_sparse_moe.experts.6.w2.weight": "model-00018-of-00019.safetensors",
710
+ "model.layers.29.block_sparse_moe.experts.6.w3.weight": "model-00018-of-00019.safetensors",
711
+ "model.layers.29.block_sparse_moe.experts.7.w1.weight": "model-00018-of-00019.safetensors",
712
+ "model.layers.29.block_sparse_moe.experts.7.w2.weight": "model-00018-of-00019.safetensors",
713
+ "model.layers.29.block_sparse_moe.experts.7.w3.weight": "model-00018-of-00019.safetensors",
714
+ "model.layers.29.block_sparse_moe.gate.weight": "model-00018-of-00019.safetensors",
715
+ "model.layers.29.input_layernorm.weight": "model-00018-of-00019.safetensors",
716
+ "model.layers.29.post_attention_layernorm.weight": "model-00018-of-00019.safetensors",
717
+ "model.layers.29.self_attn.k_proj.weight": "model-00018-of-00019.safetensors",
718
+ "model.layers.29.self_attn.o_proj.weight": "model-00018-of-00019.safetensors",
719
+ "model.layers.29.self_attn.q_proj.weight": "model-00018-of-00019.safetensors",
720
+ "model.layers.29.self_attn.v_proj.weight": "model-00018-of-00019.safetensors",
721
+ "model.layers.3.block_sparse_moe.experts.0.w1.weight": "model-00002-of-00019.safetensors",
722
+ "model.layers.3.block_sparse_moe.experts.0.w2.weight": "model-00002-of-00019.safetensors",
723
+ "model.layers.3.block_sparse_moe.experts.0.w3.weight": "model-00002-of-00019.safetensors",
724
+ "model.layers.3.block_sparse_moe.experts.1.w1.weight": "model-00002-of-00019.safetensors",
725
+ "model.layers.3.block_sparse_moe.experts.1.w2.weight": "model-00002-of-00019.safetensors",
726
+ "model.layers.3.block_sparse_moe.experts.1.w3.weight": "model-00002-of-00019.safetensors",
727
+ "model.layers.3.block_sparse_moe.experts.2.w1.weight": "model-00002-of-00019.safetensors",
728
+ "model.layers.3.block_sparse_moe.experts.2.w2.weight": "model-00003-of-00019.safetensors",
729
+ "model.layers.3.block_sparse_moe.experts.2.w3.weight": "model-00003-of-00019.safetensors",
730
+ "model.layers.3.block_sparse_moe.experts.3.w1.weight": "model-00003-of-00019.safetensors",
731
+ "model.layers.3.block_sparse_moe.experts.3.w2.weight": "model-00003-of-00019.safetensors",
732
+ "model.layers.3.block_sparse_moe.experts.3.w3.weight": "model-00003-of-00019.safetensors",
733
+ "model.layers.3.block_sparse_moe.experts.4.w1.weight": "model-00003-of-00019.safetensors",
734
+ "model.layers.3.block_sparse_moe.experts.4.w2.weight": "model-00003-of-00019.safetensors",
735
+ "model.layers.3.block_sparse_moe.experts.4.w3.weight": "model-00003-of-00019.safetensors",
736
+ "model.layers.3.block_sparse_moe.experts.5.w1.weight": "model-00003-of-00019.safetensors",
737
+ "model.layers.3.block_sparse_moe.experts.5.w2.weight": "model-00003-of-00019.safetensors",
738
+ "model.layers.3.block_sparse_moe.experts.5.w3.weight": "model-00003-of-00019.safetensors",
739
+ "model.layers.3.block_sparse_moe.experts.6.w1.weight": "model-00003-of-00019.safetensors",
740
+ "model.layers.3.block_sparse_moe.experts.6.w2.weight": "model-00003-of-00019.safetensors",
741
+ "model.layers.3.block_sparse_moe.experts.6.w3.weight": "model-00003-of-00019.safetensors",
742
+ "model.layers.3.block_sparse_moe.experts.7.w1.weight": "model-00003-of-00019.safetensors",
743
+ "model.layers.3.block_sparse_moe.experts.7.w2.weight": "model-00003-of-00019.safetensors",
744
+ "model.layers.3.block_sparse_moe.experts.7.w3.weight": "model-00003-of-00019.safetensors",
745
+ "model.layers.3.block_sparse_moe.gate.weight": "model-00002-of-00019.safetensors",
746
+ "model.layers.3.input_layernorm.weight": "model-00003-of-00019.safetensors",
747
+ "model.layers.3.post_attention_layernorm.weight": "model-00003-of-00019.safetensors",
748
+ "model.layers.3.self_attn.k_proj.weight": "model-00002-of-00019.safetensors",
749
+ "model.layers.3.self_attn.o_proj.weight": "model-00002-of-00019.safetensors",
750
+ "model.layers.3.self_attn.q_proj.weight": "model-00002-of-00019.safetensors",
751
+ "model.layers.3.self_attn.v_proj.weight": "model-00002-of-00019.safetensors",
752
+ "model.layers.30.block_sparse_moe.experts.0.w1.weight": "model-00018-of-00019.safetensors",
753
+ "model.layers.30.block_sparse_moe.experts.0.w2.weight": "model-00018-of-00019.safetensors",
754
+ "model.layers.30.block_sparse_moe.experts.0.w3.weight": "model-00018-of-00019.safetensors",
755
+ "model.layers.30.block_sparse_moe.experts.1.w1.weight": "model-00018-of-00019.safetensors",
756
+ "model.layers.30.block_sparse_moe.experts.1.w2.weight": "model-00018-of-00019.safetensors",
757
+ "model.layers.30.block_sparse_moe.experts.1.w3.weight": "model-00018-of-00019.safetensors",
758
+ "model.layers.30.block_sparse_moe.experts.2.w1.weight": "model-00018-of-00019.safetensors",
759
+ "model.layers.30.block_sparse_moe.experts.2.w2.weight": "model-00018-of-00019.safetensors",
760
+ "model.layers.30.block_sparse_moe.experts.2.w3.weight": "model-00018-of-00019.safetensors",
761
+ "model.layers.30.block_sparse_moe.experts.3.w1.weight": "model-00018-of-00019.safetensors",
762
+ "model.layers.30.block_sparse_moe.experts.3.w2.weight": "model-00018-of-00019.safetensors",
763
+ "model.layers.30.block_sparse_moe.experts.3.w3.weight": "model-00018-of-00019.safetensors",
764
+ "model.layers.30.block_sparse_moe.experts.4.w1.weight": "model-00018-of-00019.safetensors",
765
+ "model.layers.30.block_sparse_moe.experts.4.w2.weight": "model-00018-of-00019.safetensors",
766
+ "model.layers.30.block_sparse_moe.experts.4.w3.weight": "model-00018-of-00019.safetensors",
767
+ "model.layers.30.block_sparse_moe.experts.5.w1.weight": "model-00019-of-00019.safetensors",
768
+ "model.layers.30.block_sparse_moe.experts.5.w2.weight": "model-00019-of-00019.safetensors",
769
+ "model.layers.30.block_sparse_moe.experts.5.w3.weight": "model-00019-of-00019.safetensors",
770
+ "model.layers.30.block_sparse_moe.experts.6.w1.weight": "model-00019-of-00019.safetensors",
771
+ "model.layers.30.block_sparse_moe.experts.6.w2.weight": "model-00019-of-00019.safetensors",
772
+ "model.layers.30.block_sparse_moe.experts.6.w3.weight": "model-00019-of-00019.safetensors",
773
+ "model.layers.30.block_sparse_moe.experts.7.w1.weight": "model-00019-of-00019.safetensors",
774
+ "model.layers.30.block_sparse_moe.experts.7.w2.weight": "model-00019-of-00019.safetensors",
775
+ "model.layers.30.block_sparse_moe.experts.7.w3.weight": "model-00019-of-00019.safetensors",
776
+ "model.layers.30.block_sparse_moe.gate.weight": "model-00018-of-00019.safetensors",
777
+ "model.layers.30.input_layernorm.weight": "model-00019-of-00019.safetensors",
778
+ "model.layers.30.post_attention_layernorm.weight": "model-00019-of-00019.safetensors",
779
+ "model.layers.30.self_attn.k_proj.weight": "model-00018-of-00019.safetensors",
780
+ "model.layers.30.self_attn.o_proj.weight": "model-00018-of-00019.safetensors",
781
+ "model.layers.30.self_attn.q_proj.weight": "model-00018-of-00019.safetensors",
782
+ "model.layers.30.self_attn.v_proj.weight": "model-00018-of-00019.safetensors",
783
+ "model.layers.31.block_sparse_moe.experts.0.w1.weight": "model-00019-of-00019.safetensors",
784
+ "model.layers.31.block_sparse_moe.experts.0.w2.weight": "model-00019-of-00019.safetensors",
785
+ "model.layers.31.block_sparse_moe.experts.0.w3.weight": "model-00019-of-00019.safetensors",
786
+ "model.layers.31.block_sparse_moe.experts.1.w1.weight": "model-00019-of-00019.safetensors",
787
+ "model.layers.31.block_sparse_moe.experts.1.w2.weight": "model-00019-of-00019.safetensors",
788
+ "model.layers.31.block_sparse_moe.experts.1.w3.weight": "model-00019-of-00019.safetensors",
789
+ "model.layers.31.block_sparse_moe.experts.2.w1.weight": "model-00019-of-00019.safetensors",
790
+ "model.layers.31.block_sparse_moe.experts.2.w2.weight": "model-00019-of-00019.safetensors",
791
+ "model.layers.31.block_sparse_moe.experts.2.w3.weight": "model-00019-of-00019.safetensors",
792
+ "model.layers.31.block_sparse_moe.experts.3.w1.weight": "model-00019-of-00019.safetensors",
793
+ "model.layers.31.block_sparse_moe.experts.3.w2.weight": "model-00019-of-00019.safetensors",
794
+ "model.layers.31.block_sparse_moe.experts.3.w3.weight": "model-00019-of-00019.safetensors",
795
+ "model.layers.31.block_sparse_moe.experts.4.w1.weight": "model-00019-of-00019.safetensors",
796
+ "model.layers.31.block_sparse_moe.experts.4.w2.weight": "model-00019-of-00019.safetensors",
797
+ "model.layers.31.block_sparse_moe.experts.4.w3.weight": "model-00019-of-00019.safetensors",
798
+ "model.layers.31.block_sparse_moe.experts.5.w1.weight": "model-00019-of-00019.safetensors",
799
+ "model.layers.31.block_sparse_moe.experts.5.w2.weight": "model-00019-of-00019.safetensors",
800
+ "model.layers.31.block_sparse_moe.experts.5.w3.weight": "model-00019-of-00019.safetensors",
801
+ "model.layers.31.block_sparse_moe.experts.6.w1.weight": "model-00019-of-00019.safetensors",
802
+ "model.layers.31.block_sparse_moe.experts.6.w2.weight": "model-00019-of-00019.safetensors",
803
+ "model.layers.31.block_sparse_moe.experts.6.w3.weight": "model-00019-of-00019.safetensors",
804
+ "model.layers.31.block_sparse_moe.experts.7.w1.weight": "model-00019-of-00019.safetensors",
805
+ "model.layers.31.block_sparse_moe.experts.7.w2.weight": "model-00019-of-00019.safetensors",
806
+ "model.layers.31.block_sparse_moe.experts.7.w3.weight": "model-00019-of-00019.safetensors",
807
+ "model.layers.31.block_sparse_moe.gate.weight": "model-00019-of-00019.safetensors",
808
+ "model.layers.31.input_layernorm.weight": "model-00019-of-00019.safetensors",
809
+ "model.layers.31.post_attention_layernorm.weight": "model-00019-of-00019.safetensors",
810
+ "model.layers.31.self_attn.k_proj.weight": "model-00019-of-00019.safetensors",
811
+ "model.layers.31.self_attn.o_proj.weight": "model-00019-of-00019.safetensors",
812
+ "model.layers.31.self_attn.q_proj.weight": "model-00019-of-00019.safetensors",
813
+ "model.layers.31.self_attn.v_proj.weight": "model-00019-of-00019.safetensors",
814
+ "model.layers.4.block_sparse_moe.experts.0.w1.weight": "model-00003-of-00019.safetensors",
815
+ "model.layers.4.block_sparse_moe.experts.0.w2.weight": "model-00003-of-00019.safetensors",
816
+ "model.layers.4.block_sparse_moe.experts.0.w3.weight": "model-00003-of-00019.safetensors",
817
+ "model.layers.4.block_sparse_moe.experts.1.w1.weight": "model-00003-of-00019.safetensors",
818
+ "model.layers.4.block_sparse_moe.experts.1.w2.weight": "model-00003-of-00019.safetensors",
819
+ "model.layers.4.block_sparse_moe.experts.1.w3.weight": "model-00003-of-00019.safetensors",
820
+ "model.layers.4.block_sparse_moe.experts.2.w1.weight": "model-00003-of-00019.safetensors",
821
+ "model.layers.4.block_sparse_moe.experts.2.w2.weight": "model-00003-of-00019.safetensors",
822
+ "model.layers.4.block_sparse_moe.experts.2.w3.weight": "model-00003-of-00019.safetensors",
823
+ "model.layers.4.block_sparse_moe.experts.3.w1.weight": "model-00003-of-00019.safetensors",
824
+ "model.layers.4.block_sparse_moe.experts.3.w2.weight": "model-00003-of-00019.safetensors",
825
+ "model.layers.4.block_sparse_moe.experts.3.w3.weight": "model-00003-of-00019.safetensors",
826
+ "model.layers.4.block_sparse_moe.experts.4.w1.weight": "model-00003-of-00019.safetensors",
827
+ "model.layers.4.block_sparse_moe.experts.4.w2.weight": "model-00003-of-00019.safetensors",
828
+ "model.layers.4.block_sparse_moe.experts.4.w3.weight": "model-00003-of-00019.safetensors",
829
+ "model.layers.4.block_sparse_moe.experts.5.w1.weight": "model-00003-of-00019.safetensors",
830
+ "model.layers.4.block_sparse_moe.experts.5.w2.weight": "model-00003-of-00019.safetensors",
831
+ "model.layers.4.block_sparse_moe.experts.5.w3.weight": "model-00003-of-00019.safetensors",
832
+ "model.layers.4.block_sparse_moe.experts.6.w1.weight": "model-00003-of-00019.safetensors",
833
+ "model.layers.4.block_sparse_moe.experts.6.w2.weight": "model-00003-of-00019.safetensors",
834
+ "model.layers.4.block_sparse_moe.experts.6.w3.weight": "model-00003-of-00019.safetensors",
835
+ "model.layers.4.block_sparse_moe.experts.7.w1.weight": "model-00003-of-00019.safetensors",
836
+ "model.layers.4.block_sparse_moe.experts.7.w2.weight": "model-00003-of-00019.safetensors",
837
+ "model.layers.4.block_sparse_moe.experts.7.w3.weight": "model-00003-of-00019.safetensors",
838
+ "model.layers.4.block_sparse_moe.gate.weight": "model-00003-of-00019.safetensors",
839
+ "model.layers.4.input_layernorm.weight": "model-00003-of-00019.safetensors",
840
+ "model.layers.4.post_attention_layernorm.weight": "model-00003-of-00019.safetensors",
841
+ "model.layers.4.self_attn.k_proj.weight": "model-00003-of-00019.safetensors",
842
+ "model.layers.4.self_attn.o_proj.weight": "model-00003-of-00019.safetensors",
843
+ "model.layers.4.self_attn.q_proj.weight": "model-00003-of-00019.safetensors",
844
+ "model.layers.4.self_attn.v_proj.weight": "model-00003-of-00019.safetensors",
845
+ "model.layers.5.block_sparse_moe.experts.0.w1.weight": "model-00004-of-00019.safetensors",
846
+ "model.layers.5.block_sparse_moe.experts.0.w2.weight": "model-00004-of-00019.safetensors",
847
+ "model.layers.5.block_sparse_moe.experts.0.w3.weight": "model-00004-of-00019.safetensors",
848
+ "model.layers.5.block_sparse_moe.experts.1.w1.weight": "model-00004-of-00019.safetensors",
849
+ "model.layers.5.block_sparse_moe.experts.1.w2.weight": "model-00004-of-00019.safetensors",
850
+ "model.layers.5.block_sparse_moe.experts.1.w3.weight": "model-00004-of-00019.safetensors",
851
+ "model.layers.5.block_sparse_moe.experts.2.w1.weight": "model-00004-of-00019.safetensors",
852
+ "model.layers.5.block_sparse_moe.experts.2.w2.weight": "model-00004-of-00019.safetensors",
853
+ "model.layers.5.block_sparse_moe.experts.2.w3.weight": "model-00004-of-00019.safetensors",
854
+ "model.layers.5.block_sparse_moe.experts.3.w1.weight": "model-00004-of-00019.safetensors",
855
+ "model.layers.5.block_sparse_moe.experts.3.w2.weight": "model-00004-of-00019.safetensors",
856
+ "model.layers.5.block_sparse_moe.experts.3.w3.weight": "model-00004-of-00019.safetensors",
857
+ "model.layers.5.block_sparse_moe.experts.4.w1.weight": "model-00004-of-00019.safetensors",
858
+ "model.layers.5.block_sparse_moe.experts.4.w2.weight": "model-00004-of-00019.safetensors",
859
+ "model.layers.5.block_sparse_moe.experts.4.w3.weight": "model-00004-of-00019.safetensors",
860
+ "model.layers.5.block_sparse_moe.experts.5.w1.weight": "model-00004-of-00019.safetensors",
861
+ "model.layers.5.block_sparse_moe.experts.5.w2.weight": "model-00004-of-00019.safetensors",
862
+ "model.layers.5.block_sparse_moe.experts.5.w3.weight": "model-00004-of-00019.safetensors",
863
+ "model.layers.5.block_sparse_moe.experts.6.w1.weight": "model-00004-of-00019.safetensors",
864
+ "model.layers.5.block_sparse_moe.experts.6.w2.weight": "model-00004-of-00019.safetensors",
865
+ "model.layers.5.block_sparse_moe.experts.6.w3.weight": "model-00004-of-00019.safetensors",
866
+ "model.layers.5.block_sparse_moe.experts.7.w1.weight": "model-00004-of-00019.safetensors",
867
+ "model.layers.5.block_sparse_moe.experts.7.w2.weight": "model-00004-of-00019.safetensors",
868
+ "model.layers.5.block_sparse_moe.experts.7.w3.weight": "model-00004-of-00019.safetensors",
869
+ "model.layers.5.block_sparse_moe.gate.weight": "model-00003-of-00019.safetensors",
870
+ "model.layers.5.input_layernorm.weight": "model-00004-of-00019.safetensors",
871
+ "model.layers.5.post_attention_layernorm.weight": "model-00004-of-00019.safetensors",
872
+ "model.layers.5.self_attn.k_proj.weight": "model-00003-of-00019.safetensors",
873
+ "model.layers.5.self_attn.o_proj.weight": "model-00003-of-00019.safetensors",
874
+ "model.layers.5.self_attn.q_proj.weight": "model-00003-of-00019.safetensors",
875
+ "model.layers.5.self_attn.v_proj.weight": "model-00003-of-00019.safetensors",
876
+ "model.layers.6.block_sparse_moe.experts.0.w1.weight": "model-00004-of-00019.safetensors",
877
+ "model.layers.6.block_sparse_moe.experts.0.w2.weight": "model-00004-of-00019.safetensors",
878
+ "model.layers.6.block_sparse_moe.experts.0.w3.weight": "model-00004-of-00019.safetensors",
879
+ "model.layers.6.block_sparse_moe.experts.1.w1.weight": "model-00004-of-00019.safetensors",
880
+ "model.layers.6.block_sparse_moe.experts.1.w2.weight": "model-00004-of-00019.safetensors",
881
+ "model.layers.6.block_sparse_moe.experts.1.w3.weight": "model-00004-of-00019.safetensors",
882
+ "model.layers.6.block_sparse_moe.experts.2.w1.weight": "model-00004-of-00019.safetensors",
883
+ "model.layers.6.block_sparse_moe.experts.2.w2.weight": "model-00004-of-00019.safetensors",
884
+ "model.layers.6.block_sparse_moe.experts.2.w3.weight": "model-00004-of-00019.safetensors",
885
+ "model.layers.6.block_sparse_moe.experts.3.w1.weight": "model-00004-of-00019.safetensors",
886
+ "model.layers.6.block_sparse_moe.experts.3.w2.weight": "model-00004-of-00019.safetensors",
887
+ "model.layers.6.block_sparse_moe.experts.3.w3.weight": "model-00004-of-00019.safetensors",
888
+ "model.layers.6.block_sparse_moe.experts.4.w1.weight": "model-00004-of-00019.safetensors",
889
+ "model.layers.6.block_sparse_moe.experts.4.w2.weight": "model-00004-of-00019.safetensors",
890
+ "model.layers.6.block_sparse_moe.experts.4.w3.weight": "model-00004-of-00019.safetensors",
891
+ "model.layers.6.block_sparse_moe.experts.5.w1.weight": "model-00004-of-00019.safetensors",
892
+ "model.layers.6.block_sparse_moe.experts.5.w2.weight": "model-00004-of-00019.safetensors",
893
+ "model.layers.6.block_sparse_moe.experts.5.w3.weight": "model-00005-of-00019.safetensors",
894
+ "model.layers.6.block_sparse_moe.experts.6.w1.weight": "model-00005-of-00019.safetensors",
895
+ "model.layers.6.block_sparse_moe.experts.6.w2.weight": "model-00005-of-00019.safetensors",
896
+ "model.layers.6.block_sparse_moe.experts.6.w3.weight": "model-00005-of-00019.safetensors",
897
+ "model.layers.6.block_sparse_moe.experts.7.w1.weight": "model-00005-of-00019.safetensors",
898
+ "model.layers.6.block_sparse_moe.experts.7.w2.weight": "model-00005-of-00019.safetensors",
899
+ "model.layers.6.block_sparse_moe.experts.7.w3.weight": "model-00005-of-00019.safetensors",
900
+ "model.layers.6.block_sparse_moe.gate.weight": "model-00004-of-00019.safetensors",
901
+ "model.layers.6.input_layernorm.weight": "model-00005-of-00019.safetensors",
902
+ "model.layers.6.post_attention_layernorm.weight": "model-00005-of-00019.safetensors",
903
+ "model.layers.6.self_attn.k_proj.weight": "model-00004-of-00019.safetensors",
904
+ "model.layers.6.self_attn.o_proj.weight": "model-00004-of-00019.safetensors",
905
+ "model.layers.6.self_attn.q_proj.weight": "model-00004-of-00019.safetensors",
906
+ "model.layers.6.self_attn.v_proj.weight": "model-00004-of-00019.safetensors",
907
+ "model.layers.7.block_sparse_moe.experts.0.w1.weight": "model-00005-of-00019.safetensors",
908
+ "model.layers.7.block_sparse_moe.experts.0.w2.weight": "model-00005-of-00019.safetensors",
909
+ "model.layers.7.block_sparse_moe.experts.0.w3.weight": "model-00005-of-00019.safetensors",
910
+ "model.layers.7.block_sparse_moe.experts.1.w1.weight": "model-00005-of-00019.safetensors",
911
+ "model.layers.7.block_sparse_moe.experts.1.w2.weight": "model-00005-of-00019.safetensors",
912
+ "model.layers.7.block_sparse_moe.experts.1.w3.weight": "model-00005-of-00019.safetensors",
913
+ "model.layers.7.block_sparse_moe.experts.2.w1.weight": "model-00005-of-00019.safetensors",
914
+ "model.layers.7.block_sparse_moe.experts.2.w2.weight": "model-00005-of-00019.safetensors",
915
+ "model.layers.7.block_sparse_moe.experts.2.w3.weight": "model-00005-of-00019.safetensors",
916
+ "model.layers.7.block_sparse_moe.experts.3.w1.weight": "model-00005-of-00019.safetensors",
917
+ "model.layers.7.block_sparse_moe.experts.3.w2.weight": "model-00005-of-00019.safetensors",
918
+ "model.layers.7.block_sparse_moe.experts.3.w3.weight": "model-00005-of-00019.safetensors",
919
+ "model.layers.7.block_sparse_moe.experts.4.w1.weight": "model-00005-of-00019.safetensors",
920
+ "model.layers.7.block_sparse_moe.experts.4.w2.weight": "model-00005-of-00019.safetensors",
921
+ "model.layers.7.block_sparse_moe.experts.4.w3.weight": "model-00005-of-00019.safetensors",
922
+ "model.layers.7.block_sparse_moe.experts.5.w1.weight": "model-00005-of-00019.safetensors",
923
+ "model.layers.7.block_sparse_moe.experts.5.w2.weight": "model-00005-of-00019.safetensors",
924
+ "model.layers.7.block_sparse_moe.experts.5.w3.weight": "model-00005-of-00019.safetensors",
925
+ "model.layers.7.block_sparse_moe.experts.6.w1.weight": "model-00005-of-00019.safetensors",
926
+ "model.layers.7.block_sparse_moe.experts.6.w2.weight": "model-00005-of-00019.safetensors",
927
+ "model.layers.7.block_sparse_moe.experts.6.w3.weight": "model-00005-of-00019.safetensors",
928
+ "model.layers.7.block_sparse_moe.experts.7.w1.weight": "model-00005-of-00019.safetensors",
929
+ "model.layers.7.block_sparse_moe.experts.7.w2.weight": "model-00005-of-00019.safetensors",
930
+ "model.layers.7.block_sparse_moe.experts.7.w3.weight": "model-00005-of-00019.safetensors",
931
+ "model.layers.7.block_sparse_moe.gate.weight": "model-00005-of-00019.safetensors",
932
+ "model.layers.7.input_layernorm.weight": "model-00005-of-00019.safetensors",
933
+ "model.layers.7.post_attention_layernorm.weight": "model-00005-of-00019.safetensors",
934
+ "model.layers.7.self_attn.k_proj.weight": "model-00005-of-00019.safetensors",
935
+ "model.layers.7.self_attn.o_proj.weight": "model-00005-of-00019.safetensors",
936
+ "model.layers.7.self_attn.q_proj.weight": "model-00005-of-00019.safetensors",
937
+ "model.layers.7.self_attn.v_proj.weight": "model-00005-of-00019.safetensors",
938
+ "model.layers.8.block_sparse_moe.experts.0.w1.weight": "model-00005-of-00019.safetensors",
939
+ "model.layers.8.block_sparse_moe.experts.0.w2.weight": "model-00005-of-00019.safetensors",
940
+ "model.layers.8.block_sparse_moe.experts.0.w3.weight": "model-00005-of-00019.safetensors",
941
+ "model.layers.8.block_sparse_moe.experts.1.w1.weight": "model-00005-of-00019.safetensors",
942
+ "model.layers.8.block_sparse_moe.experts.1.w2.weight": "model-00005-of-00019.safetensors",
943
+ "model.layers.8.block_sparse_moe.experts.1.w3.weight": "model-00005-of-00019.safetensors",
944
+ "model.layers.8.block_sparse_moe.experts.2.w1.weight": "model-00005-of-00019.safetensors",
945
+ "model.layers.8.block_sparse_moe.experts.2.w2.weight": "model-00005-of-00019.safetensors",
946
+ "model.layers.8.block_sparse_moe.experts.2.w3.weight": "model-00005-of-00019.safetensors",
947
+ "model.layers.8.block_sparse_moe.experts.3.w1.weight": "model-00005-of-00019.safetensors",
948
+ "model.layers.8.block_sparse_moe.experts.3.w2.weight": "model-00006-of-00019.safetensors",
949
+ "model.layers.8.block_sparse_moe.experts.3.w3.weight": "model-00006-of-00019.safetensors",
950
+ "model.layers.8.block_sparse_moe.experts.4.w1.weight": "model-00006-of-00019.safetensors",
951
+ "model.layers.8.block_sparse_moe.experts.4.w2.weight": "model-00006-of-00019.safetensors",
952
+ "model.layers.8.block_sparse_moe.experts.4.w3.weight": "model-00006-of-00019.safetensors",
953
+ "model.layers.8.block_sparse_moe.experts.5.w1.weight": "model-00006-of-00019.safetensors",
954
+ "model.layers.8.block_sparse_moe.experts.5.w2.weight": "model-00006-of-00019.safetensors",
955
+ "model.layers.8.block_sparse_moe.experts.5.w3.weight": "model-00006-of-00019.safetensors",
956
+ "model.layers.8.block_sparse_moe.experts.6.w1.weight": "model-00006-of-00019.safetensors",
957
+ "model.layers.8.block_sparse_moe.experts.6.w2.weight": "model-00006-of-00019.safetensors",
958
+ "model.layers.8.block_sparse_moe.experts.6.w3.weight": "model-00006-of-00019.safetensors",
959
+ "model.layers.8.block_sparse_moe.experts.7.w1.weight": "model-00006-of-00019.safetensors",
960
+ "model.layers.8.block_sparse_moe.experts.7.w2.weight": "model-00006-of-00019.safetensors",
961
+ "model.layers.8.block_sparse_moe.experts.7.w3.weight": "model-00006-of-00019.safetensors",
962
+ "model.layers.8.block_sparse_moe.gate.weight": "model-00005-of-00019.safetensors",
963
+ "model.layers.8.input_layernorm.weight": "model-00006-of-00019.safetensors",
964
+ "model.layers.8.post_attention_layernorm.weight": "model-00006-of-00019.safetensors",
965
+ "model.layers.8.self_attn.k_proj.weight": "model-00005-of-00019.safetensors",
966
+ "model.layers.8.self_attn.o_proj.weight": "model-00005-of-00019.safetensors",
967
+ "model.layers.8.self_attn.q_proj.weight": "model-00005-of-00019.safetensors",
968
+ "model.layers.8.self_attn.v_proj.weight": "model-00005-of-00019.safetensors",
969
+ "model.layers.9.block_sparse_moe.experts.0.w1.weight": "model-00006-of-00019.safetensors",
970
+ "model.layers.9.block_sparse_moe.experts.0.w2.weight": "model-00006-of-00019.safetensors",
971
+ "model.layers.9.block_sparse_moe.experts.0.w3.weight": "model-00006-of-00019.safetensors",
972
+ "model.layers.9.block_sparse_moe.experts.1.w1.weight": "model-00006-of-00019.safetensors",
973
+ "model.layers.9.block_sparse_moe.experts.1.w2.weight": "model-00006-of-00019.safetensors",
974
+ "model.layers.9.block_sparse_moe.experts.1.w3.weight": "model-00006-of-00019.safetensors",
975
+ "model.layers.9.block_sparse_moe.experts.2.w1.weight": "model-00006-of-00019.safetensors",
976
+ "model.layers.9.block_sparse_moe.experts.2.w2.weight": "model-00006-of-00019.safetensors",
977
+ "model.layers.9.block_sparse_moe.experts.2.w3.weight": "model-00006-of-00019.safetensors",
978
+ "model.layers.9.block_sparse_moe.experts.3.w1.weight": "model-00006-of-00019.safetensors",
979
+ "model.layers.9.block_sparse_moe.experts.3.w2.weight": "model-00006-of-00019.safetensors",
980
+ "model.layers.9.block_sparse_moe.experts.3.w3.weight": "model-00006-of-00019.safetensors",
981
+ "model.layers.9.block_sparse_moe.experts.4.w1.weight": "model-00006-of-00019.safetensors",
982
+ "model.layers.9.block_sparse_moe.experts.4.w2.weight": "model-00006-of-00019.safetensors",
983
+ "model.layers.9.block_sparse_moe.experts.4.w3.weight": "model-00006-of-00019.safetensors",
984
+ "model.layers.9.block_sparse_moe.experts.5.w1.weight": "model-00006-of-00019.safetensors",
985
+ "model.layers.9.block_sparse_moe.experts.5.w2.weight": "model-00006-of-00019.safetensors",
986
+ "model.layers.9.block_sparse_moe.experts.5.w3.weight": "model-00006-of-00019.safetensors",
987
+ "model.layers.9.block_sparse_moe.experts.6.w1.weight": "model-00006-of-00019.safetensors",
988
+ "model.layers.9.block_sparse_moe.experts.6.w2.weight": "model-00006-of-00019.safetensors",
989
+ "model.layers.9.block_sparse_moe.experts.6.w3.weight": "model-00006-of-00019.safetensors",
990
+ "model.layers.9.block_sparse_moe.experts.7.w1.weight": "model-00006-of-00019.safetensors",
991
+ "model.layers.9.block_sparse_moe.experts.7.w2.weight": "model-00006-of-00019.safetensors",
992
+ "model.layers.9.block_sparse_moe.experts.7.w3.weight": "model-00006-of-00019.safetensors",
993
+ "model.layers.9.block_sparse_moe.gate.weight": "model-00006-of-00019.safetensors",
994
+ "model.layers.9.input_layernorm.weight": "model-00006-of-00019.safetensors",
995
+ "model.layers.9.post_attention_layernorm.weight": "model-00006-of-00019.safetensors",
996
+ "model.layers.9.self_attn.k_proj.weight": "model-00006-of-00019.safetensors",
997
+ "model.layers.9.self_attn.o_proj.weight": "model-00006-of-00019.safetensors",
998
+ "model.layers.9.self_attn.q_proj.weight": "model-00006-of-00019.safetensors",
999
+ "model.layers.9.self_attn.v_proj.weight": "model-00006-of-00019.safetensors",
1000
+ "model.norm.weight": "model-00019-of-00019.safetensors"
1001
+ }
1002
+ }
runs/Mar07_15-07-19_algo-1/events.out.tfevents.1709824095.algo-1.118.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45fcf48aaa892c329753904df5fbd3a1c853ccad0f4e04760b84883fc87ce6b3
3
+ size 16750
runs/Mar07_15-07-19_algo-1/events.out.tfevents.1709854397.algo-1.118.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fd2e200bf763a24bb91f6e156d3407cff48154b530f6813145454c10adb94d1
3
+ size 359
special_tokens_map.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ {
11
+ "content": "</s>",
12
+ "lstrip": false,
13
+ "normalized": false,
14
+ "rstrip": false,
15
+ "single_word": false
16
+ }
17
+ ],
18
+ "bos_token": "<s>",
19
+ "eos_token": "</s>",
20
+ "pad_token": "<unk>",
21
+ "unk_token": {
22
+ "content": "<unk>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false
27
+ }
28
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
3
+ size 493443
tokenizer_config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ }
29
+ },
30
+ "additional_special_tokens": [
31
+ "<s>",
32
+ "</s>"
33
+ ],
34
+ "bos_token": "<s>",
35
+ "chat_template": "{{ bos_token }}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if message['role'] == 'user' %}{{ '[INST] ' + message['content'] + ' [/INST]' }}{% elif message['role'] == 'assistant' %}{{ message['content'] + ' ' + eos_token + ' ' }}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}",
36
+ "clean_up_tokenization_spaces": false,
37
+ "eos_token": "</s>",
38
+ "legacy": true,
39
+ "model_max_length": 1000000000000000019884624838656,
40
+ "pad_token": "<unk>",
41
+ "sp_model_kwargs": {},
42
+ "spaces_between_special_tokens": false,
43
+ "tokenizer_class": "LlamaTokenizer",
44
+ "unk_token": "<unk>",
45
+ "use_default_system_prompt": false
46
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.5342465417401601,
4
+ "train_runtime": 29202.9895,
5
+ "train_samples": 14310,
6
+ "train_samples_per_second": 0.158,
7
+ "train_steps_per_second": 0.01
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,394 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9991356957649092,
5
+ "eval_steps": 20,
6
+ "global_step": 289,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.03,
13
+ "grad_norm": 0.3046875,
14
+ "learning_rate": 2.7586206896551727e-05,
15
+ "loss": 1.6691,
16
+ "step": 8
17
+ },
18
+ {
19
+ "epoch": 0.06,
20
+ "grad_norm": 0.193359375,
21
+ "learning_rate": 5.517241379310345e-05,
22
+ "loss": 1.5039,
23
+ "step": 16
24
+ },
25
+ {
26
+ "epoch": 0.07,
27
+ "eval_loss": 1.1767374277114868,
28
+ "eval_runtime": 1099.3683,
29
+ "eval_samples_per_second": 1.039,
30
+ "eval_steps_per_second": 0.13,
31
+ "step": 20
32
+ },
33
+ {
34
+ "epoch": 0.08,
35
+ "grad_norm": 0.1591796875,
36
+ "learning_rate": 8.275862068965517e-05,
37
+ "loss": 1.1888,
38
+ "step": 24
39
+ },
40
+ {
41
+ "epoch": 0.11,
42
+ "grad_norm": 0.09033203125,
43
+ "learning_rate": 0.0001,
44
+ "loss": 0.8397,
45
+ "step": 32
46
+ },
47
+ {
48
+ "epoch": 0.14,
49
+ "grad_norm": 0.0595703125,
50
+ "learning_rate": 0.0001,
51
+ "loss": 0.6852,
52
+ "step": 40
53
+ },
54
+ {
55
+ "epoch": 0.14,
56
+ "eval_loss": 0.6382727026939392,
57
+ "eval_runtime": 1099.2109,
58
+ "eval_samples_per_second": 1.039,
59
+ "eval_steps_per_second": 0.13,
60
+ "step": 40
61
+ },
62
+ {
63
+ "epoch": 0.17,
64
+ "grad_norm": 0.06005859375,
65
+ "learning_rate": 0.0001,
66
+ "loss": 0.6169,
67
+ "step": 48
68
+ },
69
+ {
70
+ "epoch": 0.19,
71
+ "grad_norm": 0.064453125,
72
+ "learning_rate": 0.0001,
73
+ "loss": 0.6008,
74
+ "step": 56
75
+ },
76
+ {
77
+ "epoch": 0.21,
78
+ "eval_loss": 0.5603911280632019,
79
+ "eval_runtime": 1099.0172,
80
+ "eval_samples_per_second": 1.039,
81
+ "eval_steps_per_second": 0.13,
82
+ "step": 60
83
+ },
84
+ {
85
+ "epoch": 0.22,
86
+ "grad_norm": 0.06201171875,
87
+ "learning_rate": 0.0001,
88
+ "loss": 0.5572,
89
+ "step": 64
90
+ },
91
+ {
92
+ "epoch": 0.25,
93
+ "grad_norm": 0.06884765625,
94
+ "learning_rate": 0.0001,
95
+ "loss": 0.5507,
96
+ "step": 72
97
+ },
98
+ {
99
+ "epoch": 0.28,
100
+ "grad_norm": 0.06396484375,
101
+ "learning_rate": 0.0001,
102
+ "loss": 0.5262,
103
+ "step": 80
104
+ },
105
+ {
106
+ "epoch": 0.28,
107
+ "eval_loss": 0.5174164772033691,
108
+ "eval_runtime": 1099.2276,
109
+ "eval_samples_per_second": 1.039,
110
+ "eval_steps_per_second": 0.13,
111
+ "step": 80
112
+ },
113
+ {
114
+ "epoch": 0.3,
115
+ "grad_norm": 0.06787109375,
116
+ "learning_rate": 0.0001,
117
+ "loss": 0.5071,
118
+ "step": 88
119
+ },
120
+ {
121
+ "epoch": 0.33,
122
+ "grad_norm": 0.07861328125,
123
+ "learning_rate": 0.0001,
124
+ "loss": 0.5059,
125
+ "step": 96
126
+ },
127
+ {
128
+ "epoch": 0.35,
129
+ "eval_loss": 0.48593389987945557,
130
+ "eval_runtime": 1099.1884,
131
+ "eval_samples_per_second": 1.039,
132
+ "eval_steps_per_second": 0.13,
133
+ "step": 100
134
+ },
135
+ {
136
+ "epoch": 0.36,
137
+ "grad_norm": 0.08154296875,
138
+ "learning_rate": 0.0001,
139
+ "loss": 0.4696,
140
+ "step": 104
141
+ },
142
+ {
143
+ "epoch": 0.39,
144
+ "grad_norm": 0.078125,
145
+ "learning_rate": 0.0001,
146
+ "loss": 0.4786,
147
+ "step": 112
148
+ },
149
+ {
150
+ "epoch": 0.41,
151
+ "grad_norm": 0.08154296875,
152
+ "learning_rate": 0.0001,
153
+ "loss": 0.4769,
154
+ "step": 120
155
+ },
156
+ {
157
+ "epoch": 0.41,
158
+ "eval_loss": 0.4559060037136078,
159
+ "eval_runtime": 1098.9724,
160
+ "eval_samples_per_second": 1.039,
161
+ "eval_steps_per_second": 0.13,
162
+ "step": 120
163
+ },
164
+ {
165
+ "epoch": 0.44,
166
+ "grad_norm": 0.1064453125,
167
+ "learning_rate": 0.0001,
168
+ "loss": 0.4588,
169
+ "step": 128
170
+ },
171
+ {
172
+ "epoch": 0.47,
173
+ "grad_norm": 0.10595703125,
174
+ "learning_rate": 0.0001,
175
+ "loss": 0.4399,
176
+ "step": 136
177
+ },
178
+ {
179
+ "epoch": 0.48,
180
+ "eval_loss": 0.42399317026138306,
181
+ "eval_runtime": 1099.1888,
182
+ "eval_samples_per_second": 1.039,
183
+ "eval_steps_per_second": 0.13,
184
+ "step": 140
185
+ },
186
+ {
187
+ "epoch": 0.5,
188
+ "grad_norm": 0.0966796875,
189
+ "learning_rate": 0.0001,
190
+ "loss": 0.4681,
191
+ "step": 144
192
+ },
193
+ {
194
+ "epoch": 0.53,
195
+ "grad_norm": 0.11181640625,
196
+ "learning_rate": 0.0001,
197
+ "loss": 0.4085,
198
+ "step": 152
199
+ },
200
+ {
201
+ "epoch": 0.55,
202
+ "grad_norm": 0.1005859375,
203
+ "learning_rate": 0.0001,
204
+ "loss": 0.4356,
205
+ "step": 160
206
+ },
207
+ {
208
+ "epoch": 0.55,
209
+ "eval_loss": 0.40195128321647644,
210
+ "eval_runtime": 1099.4392,
211
+ "eval_samples_per_second": 1.039,
212
+ "eval_steps_per_second": 0.13,
213
+ "step": 160
214
+ },
215
+ {
216
+ "epoch": 0.58,
217
+ "grad_norm": 0.10107421875,
218
+ "learning_rate": 0.0001,
219
+ "loss": 0.4044,
220
+ "step": 168
221
+ },
222
+ {
223
+ "epoch": 0.61,
224
+ "grad_norm": 0.11328125,
225
+ "learning_rate": 0.0001,
226
+ "loss": 0.4191,
227
+ "step": 176
228
+ },
229
+ {
230
+ "epoch": 0.62,
231
+ "eval_loss": 0.3850320875644684,
232
+ "eval_runtime": 1099.3747,
233
+ "eval_samples_per_second": 1.039,
234
+ "eval_steps_per_second": 0.13,
235
+ "step": 180
236
+ },
237
+ {
238
+ "epoch": 0.64,
239
+ "grad_norm": 0.1083984375,
240
+ "learning_rate": 0.0001,
241
+ "loss": 0.3917,
242
+ "step": 184
243
+ },
244
+ {
245
+ "epoch": 0.66,
246
+ "grad_norm": 0.1005859375,
247
+ "learning_rate": 0.0001,
248
+ "loss": 0.3642,
249
+ "step": 192
250
+ },
251
+ {
252
+ "epoch": 0.69,
253
+ "grad_norm": 0.0966796875,
254
+ "learning_rate": 0.0001,
255
+ "loss": 0.3887,
256
+ "step": 200
257
+ },
258
+ {
259
+ "epoch": 0.69,
260
+ "eval_loss": 0.37055206298828125,
261
+ "eval_runtime": 1099.0492,
262
+ "eval_samples_per_second": 1.039,
263
+ "eval_steps_per_second": 0.13,
264
+ "step": 200
265
+ },
266
+ {
267
+ "epoch": 0.72,
268
+ "grad_norm": 0.10546875,
269
+ "learning_rate": 0.0001,
270
+ "loss": 0.3842,
271
+ "step": 208
272
+ },
273
+ {
274
+ "epoch": 0.75,
275
+ "grad_norm": 0.109375,
276
+ "learning_rate": 0.0001,
277
+ "loss": 0.3826,
278
+ "step": 216
279
+ },
280
+ {
281
+ "epoch": 0.76,
282
+ "eval_loss": 0.3577556312084198,
283
+ "eval_runtime": 1099.1357,
284
+ "eval_samples_per_second": 1.039,
285
+ "eval_steps_per_second": 0.13,
286
+ "step": 220
287
+ },
288
+ {
289
+ "epoch": 0.77,
290
+ "grad_norm": 0.1103515625,
291
+ "learning_rate": 0.0001,
292
+ "loss": 0.387,
293
+ "step": 224
294
+ },
295
+ {
296
+ "epoch": 0.8,
297
+ "grad_norm": 0.109375,
298
+ "learning_rate": 0.0001,
299
+ "loss": 0.3379,
300
+ "step": 232
301
+ },
302
+ {
303
+ "epoch": 0.83,
304
+ "grad_norm": 0.09716796875,
305
+ "learning_rate": 0.0001,
306
+ "loss": 0.3563,
307
+ "step": 240
308
+ },
309
+ {
310
+ "epoch": 0.83,
311
+ "eval_loss": 0.3489949703216553,
312
+ "eval_runtime": 1099.2538,
313
+ "eval_samples_per_second": 1.039,
314
+ "eval_steps_per_second": 0.13,
315
+ "step": 240
316
+ },
317
+ {
318
+ "epoch": 0.86,
319
+ "grad_norm": 0.10107421875,
320
+ "learning_rate": 0.0001,
321
+ "loss": 0.3625,
322
+ "step": 248
323
+ },
324
+ {
325
+ "epoch": 0.89,
326
+ "grad_norm": 0.11328125,
327
+ "learning_rate": 0.0001,
328
+ "loss": 0.3224,
329
+ "step": 256
330
+ },
331
+ {
332
+ "epoch": 0.9,
333
+ "eval_loss": 0.3398669362068176,
334
+ "eval_runtime": 1099.0373,
335
+ "eval_samples_per_second": 1.039,
336
+ "eval_steps_per_second": 0.13,
337
+ "step": 260
338
+ },
339
+ {
340
+ "epoch": 0.91,
341
+ "grad_norm": 0.1015625,
342
+ "learning_rate": 0.0001,
343
+ "loss": 0.35,
344
+ "step": 264
345
+ },
346
+ {
347
+ "epoch": 0.94,
348
+ "grad_norm": 0.099609375,
349
+ "learning_rate": 0.0001,
350
+ "loss": 0.3504,
351
+ "step": 272
352
+ },
353
+ {
354
+ "epoch": 0.97,
355
+ "grad_norm": 0.10400390625,
356
+ "learning_rate": 0.0001,
357
+ "loss": 0.3285,
358
+ "step": 280
359
+ },
360
+ {
361
+ "epoch": 0.97,
362
+ "eval_loss": 0.3326619565486908,
363
+ "eval_runtime": 1099.1458,
364
+ "eval_samples_per_second": 1.039,
365
+ "eval_steps_per_second": 0.13,
366
+ "step": 280
367
+ },
368
+ {
369
+ "epoch": 1.0,
370
+ "grad_norm": 0.109375,
371
+ "learning_rate": 0.0001,
372
+ "loss": 0.3506,
373
+ "step": 288
374
+ },
375
+ {
376
+ "epoch": 1.0,
377
+ "step": 289,
378
+ "total_flos": 1.350622205624451e+18,
379
+ "train_loss": 0.5342465417401601,
380
+ "train_runtime": 29202.9895,
381
+ "train_samples_per_second": 0.158,
382
+ "train_steps_per_second": 0.01
383
+ }
384
+ ],
385
+ "logging_steps": 8,
386
+ "max_steps": 289,
387
+ "num_input_tokens_seen": 0,
388
+ "num_train_epochs": 1,
389
+ "save_steps": 500,
390
+ "total_flos": 1.350622205624451e+18,
391
+ "train_batch_size": 2,
392
+ "trial_name": null,
393
+ "trial_params": null
394
+ }