Hugging Face's logo

TheDrummer
/
Anubis-Pro-105B-v1

Safetensors
llama
Model card Files
xet
 Community
1
Sao10K commited on
Commit
cef0a08
·
verified ·
1 Parent(s): d112c72

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +1 -0
  2. config.json +38 -0
  3. generation_config.json +12 -0
  4. model-00001-of-00044.safetensors +3 -0
  5. model-00002-of-00044.safetensors +3 -0
  6. model-00003-of-00044.safetensors +3 -0
  7. model-00004-of-00044.safetensors +3 -0
  8. model-00005-of-00044.safetensors +3 -0
  9. model-00006-of-00044.safetensors +3 -0
  10. model-00007-of-00044.safetensors +3 -0
  11. model-00008-of-00044.safetensors +3 -0
  12. model-00009-of-00044.safetensors +3 -0
  13. model-00010-of-00044.safetensors +3 -0
  14. model-00011-of-00044.safetensors +3 -0
  15. model-00012-of-00044.safetensors +3 -0
  16. model-00013-of-00044.safetensors +3 -0
  17. model-00014-of-00044.safetensors +3 -0
  18. model-00015-of-00044.safetensors +3 -0
  19. model-00016-of-00044.safetensors +3 -0
  20. model-00017-of-00044.safetensors +3 -0
  21. model-00018-of-00044.safetensors +3 -0
  22. model-00019-of-00044.safetensors +3 -0
  23. model-00020-of-00044.safetensors +3 -0
  24. model-00021-of-00044.safetensors +3 -0
  25. model-00022-of-00044.safetensors +3 -0
  26. model-00023-of-00044.safetensors +3 -0
  27. model-00024-of-00044.safetensors +3 -0
  28. model-00025-of-00044.safetensors +3 -0
  29. model-00026-of-00044.safetensors +3 -0
  30. model-00027-of-00044.safetensors +3 -0
  31. model-00028-of-00044.safetensors +3 -0
  32. model-00029-of-00044.safetensors +3 -0
  33. model-00030-of-00044.safetensors +3 -0
  34. model-00031-of-00044.safetensors +3 -0
  35. model-00032-of-00044.safetensors +3 -0
  36. model-00033-of-00044.safetensors +3 -0
  37. model-00034-of-00044.safetensors +3 -0
  38. model-00035-of-00044.safetensors +3 -0
  39. model-00036-of-00044.safetensors +3 -0
  40. model-00037-of-00044.safetensors +3 -0
  41. model-00038-of-00044.safetensors +3 -0
  42. model-00039-of-00044.safetensors +3 -0
  43. model-00040-of-00044.safetensors +3 -0
  44. model-00041-of-00044.safetensors +3 -0
  45. model-00042-of-00044.safetensors +3 -0
  46. model-00043-of-00044.safetensors +3 -0
  47. model-00044-of-00044.safetensors +3 -0
  48. model.safetensors.index.json +1090 -0
  49. special_tokens_map.json +23 -0
  50. tokenizer.json +3 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "BeaverAI/Anubis-Pro-105B-v1a",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 128000,
9
+ "eos_token_id": 128009,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 8192,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 28672,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 64,
19
+ "num_hidden_layers": 120,
20
+ "num_key_value_heads": 8,
21
+ "pad_token_id": 128004,
22
+ "pretraining_tp": 1,
23
+ "rms_norm_eps": 1e-05,
24
+ "rope_scaling": {
25
+ "factor": 8.0,
26
+ "high_freq_factor": 4.0,
27
+ "low_freq_factor": 1.0,
28
+ "original_max_position_embeddings": 8192,
29
+ "rope_type": "llama3"
30
+ },
31
+ "rope_theta": 500000.0,
32
+ "tie_word_embeddings": false,
33
+ "torch_dtype": "bfloat16",
34
+ "transformers_version": "4.48.1",
35
+ "unsloth_fixed": true,
36
+ "use_cache": false,
37
+ "vocab_size": 128256
38
+ }
generation_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 128000,
4
+ "do_sample": true,
5
+ "eos_token_id": [
6
+ 128001,
7
+ 128008,
8
+ 128009
9
+ ],
10
+ "pad_token_id": 128004,
11
+ "transformers_version": "4.48.1"
12
+ }
model-00001-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:71f5c67606842833805b5269a83cb420f49f8a96a42fac204d5fd41ff0b29ed3
3
+ size 4584408808
model-00002-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:13742bd73bf2bdd1785edfd01a6b549b8db792ef6a08d623f1468db05dcba49f
3
+ size 4664167376
model-00003-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38eb004272e069ed28bb3c8b48dcc25c0d1705d2fa9c05d35ca0b9c21174f46f
3
+ size 4999711704
model-00004-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c85fa506105f559608f39b3d946c92dea7bc9d01cdd202654ccd33399ba47431
3
+ size 4966157032
model-00005-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ee6a124570fe6f03d9d3bc540c2f587ce21d2ce6c91a4ad921e77d2b76d11c3
3
+ size 4664134408
model-00006-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:871a4ec7f9bdc4f606960028d4e1f1aaaac656ef387b4fff5bb33fa26a1c31bc
3
+ size 4664167408
model-00007-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0dccfefd582c3b71ca52dd5d67f439cf7ad58acd70b12525f73affc006e2ff1a
3
+ size 4664167408
model-00008-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f68faed219a5a4da83e4a10e998342bab626414cc97fbb51de242a72e256e36f
3
+ size 4999711728
model-00009-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:84135f23a4d0409c522d22668e385eb4596e05ad602b5190d1797764418e89f0
3
+ size 4966157056
model-00010-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:82e988053dd540b1cfdfc340545f1c9e32869c802a3e2cd6e7e341da65810d4e
3
+ size 4664134408
model-00011-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6b21b4fc540a7d5aed6c85d130009f488760eed6789456292556d93c1b325c6
3
+ size 4664167408
model-00012-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efda5bee25072647b9d54d5c9ca0da10d035f39ee282fa9f3a9c4070b8edb47f
3
+ size 4664167408
model-00013-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cdce372707fd2b4a913121e19c91e5b12787c507915e6e581d348cefbfcf47ac
3
+ size 4999711728
model-00014-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f93e50daffceee52cb981d742c7e2580b0762a870c1b4520cb51f5461406ff49
3
+ size 4966157056
model-00015-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c25bf1788298bb7648597baacfea4afba49e0b1f6c8ebe6843ca4dccaaf504fd
3
+ size 4664134408
model-00016-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd57b0727c9e3e6dde429b99ef6ae0148246c36e6f76226e7185e380d0dbf67d
3
+ size 4664167408
model-00017-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3eabc9ab5796e09280f10fa016eeaafd7ec4449eec36a7f72e6b31b3d76e4703
3
+ size 4664167408
model-00018-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:abbf8bea71021bd7c82ae583bcaaf0e6979f066f1881ce105b0935b09bc38a2e
3
+ size 4999711728
model-00019-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:49168e654f26224b55f367d975859471dc5581b46b63c0cfa39d0f521a4e635e
3
+ size 4966157056
model-00020-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ddeb06f28c4f255c3ec4a0c20897b4abeaad9cd7a319a4dac8dd5829bf738d7
3
+ size 4664134408
model-00021-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:830bb09f154cf1183deab0198cdb74a17fdbed7a0fe00127d387dcafd1718f8c
3
+ size 4664167408
model-00022-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c047b4bc16e6ead0aa9a71f52060cd5823279418d499eb000ae3903ebe577996
3
+ size 4664167408
model-00023-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1ff46fd80343c80a3ed7a4faefd909ce335d29bcd496c5489a5203850ae830ec
3
+ size 4999711728
model-00024-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbd3829a4ce28b15b229fc00b0441d5736070f620256e8fd83f6f29b855daef5
3
+ size 4966157056
model-00025-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9dd3b4353ac1d3ec09aa5ac607e5c7bebdd6d9ab74a2527117558eb30a6bbf1
3
+ size 4664134408
model-00026-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f8cfce270533ea938b07813fe38310539a9902dc646c0a6c4bd87a657ac0664
3
+ size 4664167408
model-00027-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab076408872a5ae8ca03b30922e0f95ae1b2a62fecb50fedc8069c08775f0e3e
3
+ size 4664167408
model-00028-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73c8f2760028d4a8d16112bff1beabe320907338b9f42853370d6f54a895f66c
3
+ size 4999711728
model-00029-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d459c2e71ea2d63957a0a98bd4924fe0542fe32c4242b91b89cb1efdbb8d4a2
3
+ size 4966157056
model-00030-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a57486a2a9c7fc92d3b346c329c90da891dbaab9a4f0c78b6c784fd5df66ad4
3
+ size 4664134408
model-00031-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3ee812f4146ef7262d5f85014b3a21ec9d776ecce442451022a76d42e99adcd3
3
+ size 4664167408
model-00032-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfb0324072c909dd2fd3be50363c3b424b2a1900562b2667f92846de578b1d2e
3
+ size 4664167408
model-00033-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a3aa4a613f785e24e19b2fb24311deecd1516b6a8f98dca7bb9c40b891e51cc1
3
+ size 4999711728
model-00034-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b806c71076ca5372cf7d888749ee817a5da4e404b04520c9a7977ac0efc9da92
3
+ size 4966157056
model-00035-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e08b77ac576ec7dd79c3dd31687e8e8962ef83ce3b29ca48143d44806b0b8f5
3
+ size 4664134408
model-00036-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01dd15c269d0fae2a03566d53627f8a97a2e2d1e900d3badd71297b0f7a57e59
3
+ size 4664167408
model-00037-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3beae486e13bced83d13d6b9485b5f3ece18cf4b7b1754384c587d476d1177c1
3
+ size 4664167432
model-00038-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f30405afd745646657af667b59a50643c2ea7cdbcc11828582115517f17252b8
3
+ size 4999711752
model-00039-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe555277dafaad29d261322bbd6a608f38771a162201127d64001ce1d3e54771
3
+ size 4966157080
model-00040-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0024f333671bf150ddefde9cafed2fd737cfdc8b4f52e70101bb7ee6405cde5
3
+ size 4664134432
model-00041-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c6c13c592dad989b2adb830bc53185417316e71b0659bc4e8aa18be1e818697
3
+ size 4664167432
model-00042-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a218904bc3d4265b1492c52f164c0a46e542a2dccb812e52af4e323ca468e3c
3
+ size 4664167432
model-00043-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a41142fa72f17c04b8e7ad921848eda21c890ff3ea70c7aa7e40e0d34c1fe7cb
3
+ size 4999711752
model-00044-of-00044.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b07e955cec733b56c50a6332f85e40af3636ca4bd25e3a2a7b23998c49d4159b
3
+ size 3644900248
model.safetensors.index.json ADDED
@@ -0,0 +1,1090 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 209559764992
4
+ },
5
+ "weight_map": {
6
+ "lm_head.weight": "model-00044-of-00044.safetensors",
7
+ "model.embed_tokens.weight": "model-00001-of-00044.safetensors",
8
+ "model.layers.0.input_layernorm.weight": "model-00001-of-00044.safetensors",
9
+ "model.layers.0.mlp.down_proj.weight": "model-00001-of-00044.safetensors",
10
+ "model.layers.0.mlp.gate_proj.weight": "model-00001-of-00044.safetensors",
11
+ "model.layers.0.mlp.up_proj.weight": "model-00001-of-00044.safetensors",
12
+ "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00044.safetensors",
13
+ "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00044.safetensors",
14
+ "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00044.safetensors",
15
+ "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00044.safetensors",
16
+ "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00044.safetensors",
17
+ "model.layers.1.input_layernorm.weight": "model-00002-of-00044.safetensors",
18
+ "model.layers.1.mlp.down_proj.weight": "model-00002-of-00044.safetensors",
19
+ "model.layers.1.mlp.gate_proj.weight": "model-00001-of-00044.safetensors",
20
+ "model.layers.1.mlp.up_proj.weight": "model-00002-of-00044.safetensors",
21
+ "model.layers.1.post_attention_layernorm.weight": "model-00002-of-00044.safetensors",
22
+ "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00044.safetensors",
23
+ "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00044.safetensors",
24
+ "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00044.safetensors",
25
+ "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00044.safetensors",
26
+ "model.layers.10.input_layernorm.weight": "model-00005-of-00044.safetensors",
27
+ "model.layers.10.mlp.down_proj.weight": "model-00005-of-00044.safetensors",
28
+ "model.layers.10.mlp.gate_proj.weight": "model-00005-of-00044.safetensors",
29
+ "model.layers.10.mlp.up_proj.weight": "model-00005-of-00044.safetensors",
30
+ "model.layers.10.post_attention_layernorm.weight": "model-00005-of-00044.safetensors",
31
+ "model.layers.10.self_attn.k_proj.weight": "model-00005-of-00044.safetensors",
32
+ "model.layers.10.self_attn.o_proj.weight": "model-00005-of-00044.safetensors",
33
+ "model.layers.10.self_attn.q_proj.weight": "model-00005-of-00044.safetensors",
34
+ "model.layers.10.self_attn.v_proj.weight": "model-00005-of-00044.safetensors",
35
+ "model.layers.100.input_layernorm.weight": "model-00037-of-00044.safetensors",
36
+ "model.layers.100.mlp.down_proj.weight": "model-00037-of-00044.safetensors",
37
+ "model.layers.100.mlp.gate_proj.weight": "model-00037-of-00044.safetensors",
38
+ "model.layers.100.mlp.up_proj.weight": "model-00037-of-00044.safetensors",
39
+ "model.layers.100.post_attention_layernorm.weight": "model-00037-of-00044.safetensors",
40
+ "model.layers.100.self_attn.k_proj.weight": "model-00037-of-00044.safetensors",
41
+ "model.layers.100.self_attn.o_proj.weight": "model-00037-of-00044.safetensors",
42
+ "model.layers.100.self_attn.q_proj.weight": "model-00037-of-00044.safetensors",
43
+ "model.layers.100.self_attn.v_proj.weight": "model-00037-of-00044.safetensors",
44
+ "model.layers.101.input_layernorm.weight": "model-00037-of-00044.safetensors",
45
+ "model.layers.101.mlp.down_proj.weight": "model-00037-of-00044.safetensors",
46
+ "model.layers.101.mlp.gate_proj.weight": "model-00037-of-00044.safetensors",
47
+ "model.layers.101.mlp.up_proj.weight": "model-00037-of-00044.safetensors",
48
+ "model.layers.101.post_attention_layernorm.weight": "model-00037-of-00044.safetensors",
49
+ "model.layers.101.self_attn.k_proj.weight": "model-00037-of-00044.safetensors",
50
+ "model.layers.101.self_attn.o_proj.weight": "model-00037-of-00044.safetensors",
51
+ "model.layers.101.self_attn.q_proj.weight": "model-00037-of-00044.safetensors",
52
+ "model.layers.101.self_attn.v_proj.weight": "model-00037-of-00044.safetensors",
53
+ "model.layers.102.input_layernorm.weight": "model-00038-of-00044.safetensors",
54
+ "model.layers.102.mlp.down_proj.weight": "model-00038-of-00044.safetensors",
55
+ "model.layers.102.mlp.gate_proj.weight": "model-00038-of-00044.safetensors",
56
+ "model.layers.102.mlp.up_proj.weight": "model-00038-of-00044.safetensors",
57
+ "model.layers.102.post_attention_layernorm.weight": "model-00038-of-00044.safetensors",
58
+ "model.layers.102.self_attn.k_proj.weight": "model-00037-of-00044.safetensors",
59
+ "model.layers.102.self_attn.o_proj.weight": "model-00037-of-00044.safetensors",
60
+ "model.layers.102.self_attn.q_proj.weight": "model-00037-of-00044.safetensors",
61
+ "model.layers.102.self_attn.v_proj.weight": "model-00037-of-00044.safetensors",
62
+ "model.layers.103.input_layernorm.weight": "model-00038-of-00044.safetensors",
63
+ "model.layers.103.mlp.down_proj.weight": "model-00038-of-00044.safetensors",
64
+ "model.layers.103.mlp.gate_proj.weight": "model-00038-of-00044.safetensors",
65
+ "model.layers.103.mlp.up_proj.weight": "model-00038-of-00044.safetensors",
66
+ "model.layers.103.post_attention_layernorm.weight": "model-00038-of-00044.safetensors",
67
+ "model.layers.103.self_attn.k_proj.weight": "model-00038-of-00044.safetensors",
68
+ "model.layers.103.self_attn.o_proj.weight": "model-00038-of-00044.safetensors",
69
+ "model.layers.103.self_attn.q_proj.weight": "model-00038-of-00044.safetensors",
70
+ "model.layers.103.self_attn.v_proj.weight": "model-00038-of-00044.safetensors",
71
+ "model.layers.104.input_layernorm.weight": "model-00038-of-00044.safetensors",
72
+ "model.layers.104.mlp.down_proj.weight": "model-00038-of-00044.safetensors",
73
+ "model.layers.104.mlp.gate_proj.weight": "model-00038-of-00044.safetensors",
74
+ "model.layers.104.mlp.up_proj.weight": "model-00038-of-00044.safetensors",
75
+ "model.layers.104.post_attention_layernorm.weight": "model-00038-of-00044.safetensors",
76
+ "model.layers.104.self_attn.k_proj.weight": "model-00038-of-00044.safetensors",
77
+ "model.layers.104.self_attn.o_proj.weight": "model-00038-of-00044.safetensors",
78
+ "model.layers.104.self_attn.q_proj.weight": "model-00038-of-00044.safetensors",
79
+ "model.layers.104.self_attn.v_proj.weight": "model-00038-of-00044.safetensors",
80
+ "model.layers.105.input_layernorm.weight": "model-00039-of-00044.safetensors",
81
+ "model.layers.105.mlp.down_proj.weight": "model-00039-of-00044.safetensors",
82
+ "model.layers.105.mlp.gate_proj.weight": "model-00039-of-00044.safetensors",
83
+ "model.layers.105.mlp.up_proj.weight": "model-00039-of-00044.safetensors",
84
+ "model.layers.105.post_attention_layernorm.weight": "model-00039-of-00044.safetensors",
85
+ "model.layers.105.self_attn.k_proj.weight": "model-00038-of-00044.safetensors",
86
+ "model.layers.105.self_attn.o_proj.weight": "model-00039-of-00044.safetensors",
87
+ "model.layers.105.self_attn.q_proj.weight": "model-00038-of-00044.safetensors",
88
+ "model.layers.105.self_attn.v_proj.weight": "model-00038-of-00044.safetensors",
89
+ "model.layers.106.input_layernorm.weight": "model-00039-of-00044.safetensors",
90
+ "model.layers.106.mlp.down_proj.weight": "model-00039-of-00044.safetensors",
91
+ "model.layers.106.mlp.gate_proj.weight": "model-00039-of-00044.safetensors",
92
+ "model.layers.106.mlp.up_proj.weight": "model-00039-of-00044.safetensors",
93
+ "model.layers.106.post_attention_layernorm.weight": "model-00039-of-00044.safetensors",
94
+ "model.layers.106.self_attn.k_proj.weight": "model-00039-of-00044.safetensors",
95
+ "model.layers.106.self_attn.o_proj.weight": "model-00039-of-00044.safetensors",
96
+ "model.layers.106.self_attn.q_proj.weight": "model-00039-of-00044.safetensors",
97
+ "model.layers.106.self_attn.v_proj.weight": "model-00039-of-00044.safetensors",
98
+ "model.layers.107.input_layernorm.weight": "model-00039-of-00044.safetensors",
99
+ "model.layers.107.mlp.down_proj.weight": "model-00039-of-00044.safetensors",
100
+ "model.layers.107.mlp.gate_proj.weight": "model-00039-of-00044.safetensors",
101
+ "model.layers.107.mlp.up_proj.weight": "model-00039-of-00044.safetensors",
102
+ "model.layers.107.post_attention_layernorm.weight": "model-00039-of-00044.safetensors",
103
+ "model.layers.107.self_attn.k_proj.weight": "model-00039-of-00044.safetensors",
104
+ "model.layers.107.self_attn.o_proj.weight": "model-00039-of-00044.safetensors",
105
+ "model.layers.107.self_attn.q_proj.weight": "model-00039-of-00044.safetensors",
106
+ "model.layers.107.self_attn.v_proj.weight": "model-00039-of-00044.safetensors",
107
+ "model.layers.108.input_layernorm.weight": "model-00040-of-00044.safetensors",
108
+ "model.layers.108.mlp.down_proj.weight": "model-00040-of-00044.safetensors",
109
+ "model.layers.108.mlp.gate_proj.weight": "model-00040-of-00044.safetensors",
110
+ "model.layers.108.mlp.up_proj.weight": "model-00040-of-00044.safetensors",
111
+ "model.layers.108.post_attention_layernorm.weight": "model-00040-of-00044.safetensors",
112
+ "model.layers.108.self_attn.k_proj.weight": "model-00040-of-00044.safetensors",
113
+ "model.layers.108.self_attn.o_proj.weight": "model-00040-of-00044.safetensors",
114
+ "model.layers.108.self_attn.q_proj.weight": "model-00040-of-00044.safetensors",
115
+ "model.layers.108.self_attn.v_proj.weight": "model-00040-of-00044.safetensors",
116
+ "model.layers.109.input_layernorm.weight": "model-00040-of-00044.safetensors",
117
+ "model.layers.109.mlp.down_proj.weight": "model-00040-of-00044.safetensors",
118
+ "model.layers.109.mlp.gate_proj.weight": "model-00040-of-00044.safetensors",
119
+ "model.layers.109.mlp.up_proj.weight": "model-00040-of-00044.safetensors",
120
+ "model.layers.109.post_attention_layernorm.weight": "model-00040-of-00044.safetensors",
121
+ "model.layers.109.self_attn.k_proj.weight": "model-00040-of-00044.safetensors",
122
+ "model.layers.109.self_attn.o_proj.weight": "model-00040-of-00044.safetensors",
123
+ "model.layers.109.self_attn.q_proj.weight": "model-00040-of-00044.safetensors",
124
+ "model.layers.109.self_attn.v_proj.weight": "model-00040-of-00044.safetensors",
125
+ "model.layers.11.input_layernorm.weight": "model-00005-of-00044.safetensors",
126
+ "model.layers.11.mlp.down_proj.weight": "model-00005-of-00044.safetensors",
127
+ "model.layers.11.mlp.gate_proj.weight": "model-00005-of-00044.safetensors",
128
+ "model.layers.11.mlp.up_proj.weight": "model-00005-of-00044.safetensors",
129
+ "model.layers.11.post_attention_layernorm.weight": "model-00005-of-00044.safetensors",
130
+ "model.layers.11.self_attn.k_proj.weight": "model-00005-of-00044.safetensors",
131
+ "model.layers.11.self_attn.o_proj.weight": "model-00005-of-00044.safetensors",
132
+ "model.layers.11.self_attn.q_proj.weight": "model-00005-of-00044.safetensors",
133
+ "model.layers.11.self_attn.v_proj.weight": "model-00005-of-00044.safetensors",
134
+ "model.layers.110.input_layernorm.weight": "model-00041-of-00044.safetensors",
135
+ "model.layers.110.mlp.down_proj.weight": "model-00041-of-00044.safetensors",
136
+ "model.layers.110.mlp.gate_proj.weight": "model-00040-of-00044.safetensors",
137
+ "model.layers.110.mlp.up_proj.weight": "model-00040-of-00044.safetensors",
138
+ "model.layers.110.post_attention_layernorm.weight": "model-00041-of-00044.safetensors",
139
+ "model.layers.110.self_attn.k_proj.weight": "model-00040-of-00044.safetensors",
140
+ "model.layers.110.self_attn.o_proj.weight": "model-00040-of-00044.safetensors",
141
+ "model.layers.110.self_attn.q_proj.weight": "model-00040-of-00044.safetensors",
142
+ "model.layers.110.self_attn.v_proj.weight": "model-00040-of-00044.safetensors",
143
+ "model.layers.111.input_layernorm.weight": "model-00041-of-00044.safetensors",
144
+ "model.layers.111.mlp.down_proj.weight": "model-00041-of-00044.safetensors",
145
+ "model.layers.111.mlp.gate_proj.weight": "model-00041-of-00044.safetensors",
146
+ "model.layers.111.mlp.up_proj.weight": "model-00041-of-00044.safetensors",
147
+ "model.layers.111.post_attention_layernorm.weight": "model-00041-of-00044.safetensors",
148
+ "model.layers.111.self_attn.k_proj.weight": "model-00041-of-00044.safetensors",
149
+ "model.layers.111.self_attn.o_proj.weight": "model-00041-of-00044.safetensors",
150
+ "model.layers.111.self_attn.q_proj.weight": "model-00041-of-00044.safetensors",
151
+ "model.layers.111.self_attn.v_proj.weight": "model-00041-of-00044.safetensors",
152
+ "model.layers.112.input_layernorm.weight": "model-00041-of-00044.safetensors",
153
+ "model.layers.112.mlp.down_proj.weight": "model-00041-of-00044.safetensors",
154
+ "model.layers.112.mlp.gate_proj.weight": "model-00041-of-00044.safetensors",
155
+ "model.layers.112.mlp.up_proj.weight": "model-00041-of-00044.safetensors",
156
+ "model.layers.112.post_attention_layernorm.weight": "model-00041-of-00044.safetensors",
157
+ "model.layers.112.self_attn.k_proj.weight": "model-00041-of-00044.safetensors",
158
+ "model.layers.112.self_attn.o_proj.weight": "model-00041-of-00044.safetensors",
159
+ "model.layers.112.self_attn.q_proj.weight": "model-00041-of-00044.safetensors",
160
+ "model.layers.112.self_attn.v_proj.weight": "model-00041-of-00044.safetensors",
161
+ "model.layers.113.input_layernorm.weight": "model-00042-of-00044.safetensors",
162
+ "model.layers.113.mlp.down_proj.weight": "model-00042-of-00044.safetensors",
163
+ "model.layers.113.mlp.gate_proj.weight": "model-00041-of-00044.safetensors",
164
+ "model.layers.113.mlp.up_proj.weight": "model-00042-of-00044.safetensors",
165
+ "model.layers.113.post_attention_layernorm.weight": "model-00042-of-00044.safetensors",
166
+ "model.layers.113.self_attn.k_proj.weight": "model-00041-of-00044.safetensors",
167
+ "model.layers.113.self_attn.o_proj.weight": "model-00041-of-00044.safetensors",
168
+ "model.layers.113.self_attn.q_proj.weight": "model-00041-of-00044.safetensors",
169
+ "model.layers.113.self_attn.v_proj.weight": "model-00041-of-00044.safetensors",
170
+ "model.layers.114.input_layernorm.weight": "model-00042-of-00044.safetensors",
171
+ "model.layers.114.mlp.down_proj.weight": "model-00042-of-00044.safetensors",
172
+ "model.layers.114.mlp.gate_proj.weight": "model-00042-of-00044.safetensors",
173
+ "model.layers.114.mlp.up_proj.weight": "model-00042-of-00044.safetensors",
174
+ "model.layers.114.post_attention_layernorm.weight": "model-00042-of-00044.safetensors",
175
+ "model.layers.114.self_attn.k_proj.weight": "model-00042-of-00044.safetensors",
176
+ "model.layers.114.self_attn.o_proj.weight": "model-00042-of-00044.safetensors",
177
+ "model.layers.114.self_attn.q_proj.weight": "model-00042-of-00044.safetensors",
178
+ "model.layers.114.self_attn.v_proj.weight": "model-00042-of-00044.safetensors",
179
+ "model.layers.115.input_layernorm.weight": "model-00042-of-00044.safetensors",
180
+ "model.layers.115.mlp.down_proj.weight": "model-00042-of-00044.safetensors",
181
+ "model.layers.115.mlp.gate_proj.weight": "model-00042-of-00044.safetensors",
182
+ "model.layers.115.mlp.up_proj.weight": "model-00042-of-00044.safetensors",
183
+ "model.layers.115.post_attention_layernorm.weight": "model-00042-of-00044.safetensors",
184
+ "model.layers.115.self_attn.k_proj.weight": "model-00042-of-00044.safetensors",
185
+ "model.layers.115.self_attn.o_proj.weight": "model-00042-of-00044.safetensors",
186
+ "model.layers.115.self_attn.q_proj.weight": "model-00042-of-00044.safetensors",
187
+ "model.layers.115.self_attn.v_proj.weight": "model-00042-of-00044.safetensors",
188
+ "model.layers.116.input_layernorm.weight": "model-00043-of-00044.safetensors",
189
+ "model.layers.116.mlp.down_proj.weight": "model-00043-of-00044.safetensors",
190
+ "model.layers.116.mlp.gate_proj.weight": "model-00043-of-00044.safetensors",
191
+ "model.layers.116.mlp.up_proj.weight": "model-00043-of-00044.safetensors",
192
+ "model.layers.116.post_attention_layernorm.weight": "model-00043-of-00044.safetensors",
193
+ "model.layers.116.self_attn.k_proj.weight": "model-00042-of-00044.safetensors",
194
+ "model.layers.116.self_attn.o_proj.weight": "model-00042-of-00044.safetensors",
195
+ "model.layers.116.self_attn.q_proj.weight": "model-00042-of-00044.safetensors",
196
+ "model.layers.116.self_attn.v_proj.weight": "model-00042-of-00044.safetensors",
197
+ "model.layers.117.input_layernorm.weight": "model-00043-of-00044.safetensors",
198
+ "model.layers.117.mlp.down_proj.weight": "model-00043-of-00044.safetensors",
199
+ "model.layers.117.mlp.gate_proj.weight": "model-00043-of-00044.safetensors",
200
+ "model.layers.117.mlp.up_proj.weight": "model-00043-of-00044.safetensors",
201
+ "model.layers.117.post_attention_layernorm.weight": "model-00043-of-00044.safetensors",
202
+ "model.layers.117.self_attn.k_proj.weight": "model-00043-of-00044.safetensors",
203
+ "model.layers.117.self_attn.o_proj.weight": "model-00043-of-00044.safetensors",
204
+ "model.layers.117.self_attn.q_proj.weight": "model-00043-of-00044.safetensors",
205
+ "model.layers.117.self_attn.v_proj.weight": "model-00043-of-00044.safetensors",
206
+ "model.layers.118.input_layernorm.weight": "model-00043-of-00044.safetensors",
207
+ "model.layers.118.mlp.down_proj.weight": "model-00043-of-00044.safetensors",
208
+ "model.layers.118.mlp.gate_proj.weight": "model-00043-of-00044.safetensors",
209
+ "model.layers.118.mlp.up_proj.weight": "model-00043-of-00044.safetensors",
210
+ "model.layers.118.post_attention_layernorm.weight": "model-00043-of-00044.safetensors",
211
+ "model.layers.118.self_attn.k_proj.weight": "model-00043-of-00044.safetensors",
212
+ "model.layers.118.self_attn.o_proj.weight": "model-00043-of-00044.safetensors",
213
+ "model.layers.118.self_attn.q_proj.weight": "model-00043-of-00044.safetensors",
214
+ "model.layers.118.self_attn.v_proj.weight": "model-00043-of-00044.safetensors",
215
+ "model.layers.119.input_layernorm.weight": "model-00044-of-00044.safetensors",
216
+ "model.layers.119.mlp.down_proj.weight": "model-00044-of-00044.safetensors",
217
+ "model.layers.119.mlp.gate_proj.weight": "model-00044-of-00044.safetensors",
218
+ "model.layers.119.mlp.up_proj.weight": "model-00044-of-00044.safetensors",
219
+ "model.layers.119.post_attention_layernorm.weight": "model-00044-of-00044.safetensors",
220
+ "model.layers.119.self_attn.k_proj.weight": "model-00043-of-00044.safetensors",
221
+ "model.layers.119.self_attn.o_proj.weight": "model-00044-of-00044.safetensors",
222
+ "model.layers.119.self_attn.q_proj.weight": "model-00043-of-00044.safetensors",
223
+ "model.layers.119.self_attn.v_proj.weight": "model-00043-of-00044.safetensors",
224
+ "model.layers.12.input_layernorm.weight": "model-00006-of-00044.safetensors",
225
+ "model.layers.12.mlp.down_proj.weight": "model-00006-of-00044.safetensors",
226
+ "model.layers.12.mlp.gate_proj.weight": "model-00005-of-00044.safetensors",
227
+ "model.layers.12.mlp.up_proj.weight": "model-00005-of-00044.safetensors",
228
+ "model.layers.12.post_attention_layernorm.weight": "model-00006-of-00044.safetensors",
229
+ "model.layers.12.self_attn.k_proj.weight": "model-00005-of-00044.safetensors",
230
+ "model.layers.12.self_attn.o_proj.weight": "model-00005-of-00044.safetensors",
231
+ "model.layers.12.self_attn.q_proj.weight": "model-00005-of-00044.safetensors",
232
+ "model.layers.12.self_attn.v_proj.weight": "model-00005-of-00044.safetensors",
233
+ "model.layers.13.input_layernorm.weight": "model-00006-of-00044.safetensors",
234
+ "model.layers.13.mlp.down_proj.weight": "model-00006-of-00044.safetensors",
235
+ "model.layers.13.mlp.gate_proj.weight": "model-00006-of-00044.safetensors",
236
+ "model.layers.13.mlp.up_proj.weight": "model-00006-of-00044.safetensors",
237
+ "model.layers.13.post_attention_layernorm.weight": "model-00006-of-00044.safetensors",
238
+ "model.layers.13.self_attn.k_proj.weight": "model-00006-of-00044.safetensors",
239
+ "model.layers.13.self_attn.o_proj.weight": "model-00006-of-00044.safetensors",
240
+ "model.layers.13.self_attn.q_proj.weight": "model-00006-of-00044.safetensors",
241
+ "model.layers.13.self_attn.v_proj.weight": "model-00006-of-00044.safetensors",
242
+ "model.layers.14.input_layernorm.weight": "model-00006-of-00044.safetensors",
243
+ "model.layers.14.mlp.down_proj.weight": "model-00006-of-00044.safetensors",
244
+ "model.layers.14.mlp.gate_proj.weight": "model-00006-of-00044.safetensors",
245
+ "model.layers.14.mlp.up_proj.weight": "model-00006-of-00044.safetensors",
246
+ "model.layers.14.post_attention_layernorm.weight": "model-00006-of-00044.safetensors",
247
+ "model.layers.14.self_attn.k_proj.weight": "model-00006-of-00044.safetensors",
248
+ "model.layers.14.self_attn.o_proj.weight": "model-00006-of-00044.safetensors",
249
+ "model.layers.14.self_attn.q_proj.weight": "model-00006-of-00044.safetensors",
250
+ "model.layers.14.self_attn.v_proj.weight": "model-00006-of-00044.safetensors",
251
+ "model.layers.15.input_layernorm.weight": "model-00007-of-00044.safetensors",
252
+ "model.layers.15.mlp.down_proj.weight": "model-00007-of-00044.safetensors",
253
+ "model.layers.15.mlp.gate_proj.weight": "model-00006-of-00044.safetensors",
254
+ "model.layers.15.mlp.up_proj.weight": "model-00007-of-00044.safetensors",
255
+ "model.layers.15.post_attention_layernorm.weight": "model-00007-of-00044.safetensors",
256
+ "model.layers.15.self_attn.k_proj.weight": "model-00006-of-00044.safetensors",
257
+ "model.layers.15.self_attn.o_proj.weight": "model-00006-of-00044.safetensors",
258
+ "model.layers.15.self_attn.q_proj.weight": "model-00006-of-00044.safetensors",
259
+ "model.layers.15.self_attn.v_proj.weight": "model-00006-of-00044.safetensors",
260
+ "model.layers.16.input_layernorm.weight": "model-00007-of-00044.safetensors",
261
+ "model.layers.16.mlp.down_proj.weight": "model-00007-of-00044.safetensors",
262
+ "model.layers.16.mlp.gate_proj.weight": "model-00007-of-00044.safetensors",
263
+ "model.layers.16.mlp.up_proj.weight": "model-00007-of-00044.safetensors",
264
+ "model.layers.16.post_attention_layernorm.weight": "model-00007-of-00044.safetensors",
265
+ "model.layers.16.self_attn.k_proj.weight": "model-00007-of-00044.safetensors",
266
+ "model.layers.16.self_attn.o_proj.weight": "model-00007-of-00044.safetensors",
267
+ "model.layers.16.self_attn.q_proj.weight": "model-00007-of-00044.safetensors",
268
+ "model.layers.16.self_attn.v_proj.weight": "model-00007-of-00044.safetensors",
269
+ "model.layers.17.input_layernorm.weight": "model-00007-of-00044.safetensors",
270
+ "model.layers.17.mlp.down_proj.weight": "model-00007-of-00044.safetensors",
271
+ "model.layers.17.mlp.gate_proj.weight": "model-00007-of-00044.safetensors",
272
+ "model.layers.17.mlp.up_proj.weight": "model-00007-of-00044.safetensors",
273
+ "model.layers.17.post_attention_layernorm.weight": "model-00007-of-00044.safetensors",
274
+ "model.layers.17.self_attn.k_proj.weight": "model-00007-of-00044.safetensors",
275
+ "model.layers.17.self_attn.o_proj.weight": "model-00007-of-00044.safetensors",
276
+ "model.layers.17.self_attn.q_proj.weight": "model-00007-of-00044.safetensors",
277
+ "model.layers.17.self_attn.v_proj.weight": "model-00007-of-00044.safetensors",
278
+ "model.layers.18.input_layernorm.weight": "model-00008-of-00044.safetensors",
279
+ "model.layers.18.mlp.down_proj.weight": "model-00008-of-00044.safetensors",
280
+ "model.layers.18.mlp.gate_proj.weight": "model-00008-of-00044.safetensors",
281
+ "model.layers.18.mlp.up_proj.weight": "model-00008-of-00044.safetensors",
282
+ "model.layers.18.post_attention_layernorm.weight": "model-00008-of-00044.safetensors",
283
+ "model.layers.18.self_attn.k_proj.weight": "model-00007-of-00044.safetensors",
284
+ "model.layers.18.self_attn.o_proj.weight": "model-00007-of-00044.safetensors",
285
+ "model.layers.18.self_attn.q_proj.weight": "model-00007-of-00044.safetensors",
286
+ "model.layers.18.self_attn.v_proj.weight": "model-00007-of-00044.safetensors",
287
+ "model.layers.19.input_layernorm.weight": "model-00008-of-00044.safetensors",
288
+ "model.layers.19.mlp.down_proj.weight": "model-00008-of-00044.safetensors",
289
+ "model.layers.19.mlp.gate_proj.weight": "model-00008-of-00044.safetensors",
290
+ "model.layers.19.mlp.up_proj.weight": "model-00008-of-00044.safetensors",
291
+ "model.layers.19.post_attention_layernorm.weight": "model-00008-of-00044.safetensors",
292
+ "model.layers.19.self_attn.k_proj.weight": "model-00008-of-00044.safetensors",
293
+ "model.layers.19.self_attn.o_proj.weight": "model-00008-of-00044.safetensors",
294
+ "model.layers.19.self_attn.q_proj.weight": "model-00008-of-00044.safetensors",
295
+ "model.layers.19.self_attn.v_proj.weight": "model-00008-of-00044.safetensors",
296
+ "model.layers.2.input_layernorm.weight": "model-00002-of-00044.safetensors",
297
+ "model.layers.2.mlp.down_proj.weight": "model-00002-of-00044.safetensors",
298
+ "model.layers.2.mlp.gate_proj.weight": "model-00002-of-00044.safetensors",
299
+ "model.layers.2.mlp.up_proj.weight": "model-00002-of-00044.safetensors",
300
+ "model.layers.2.post_attention_layernorm.weight": "model-00002-of-00044.safetensors",
301
+ "model.layers.2.self_attn.k_proj.weight": "model-00002-of-00044.safetensors",
302
+ "model.layers.2.self_attn.o_proj.weight": "model-00002-of-00044.safetensors",
303
+ "model.layers.2.self_attn.q_proj.weight": "model-00002-of-00044.safetensors",
304
+ "model.layers.2.self_attn.v_proj.weight": "model-00002-of-00044.safetensors",
305
+ "model.layers.20.input_layernorm.weight": "model-00008-of-00044.safetensors",
306
+ "model.layers.20.mlp.down_proj.weight": "model-00008-of-00044.safetensors",
307
+ "model.layers.20.mlp.gate_proj.weight": "model-00008-of-00044.safetensors",
308
+ "model.layers.20.mlp.up_proj.weight": "model-00008-of-00044.safetensors",
309
+ "model.layers.20.post_attention_layernorm.weight": "model-00008-of-00044.safetensors",
310
+ "model.layers.20.self_attn.k_proj.weight": "model-00008-of-00044.safetensors",
311
+ "model.layers.20.self_attn.o_proj.weight": "model-00008-of-00044.safetensors",
312
+ "model.layers.20.self_attn.q_proj.weight": "model-00008-of-00044.safetensors",
313
+ "model.layers.20.self_attn.v_proj.weight": "model-00008-of-00044.safetensors",
314
+ "model.layers.21.input_layernorm.weight": "model-00009-of-00044.safetensors",
315
+ "model.layers.21.mlp.down_proj.weight": "model-00009-of-00044.safetensors",
316
+ "model.layers.21.mlp.gate_proj.weight": "model-00009-of-00044.safetensors",
317
+ "model.layers.21.mlp.up_proj.weight": "model-00009-of-00044.safetensors",
318
+ "model.layers.21.post_attention_layernorm.weight": "model-00009-of-00044.safetensors",
319
+ "model.layers.21.self_attn.k_proj.weight": "model-00008-of-00044.safetensors",
320
+ "model.layers.21.self_attn.o_proj.weight": "model-00009-of-00044.safetensors",
321
+ "model.layers.21.self_attn.q_proj.weight": "model-00008-of-00044.safetensors",
322
+ "model.layers.21.self_attn.v_proj.weight": "model-00008-of-00044.safetensors",
323
+ "model.layers.22.input_layernorm.weight": "model-00009-of-00044.safetensors",
324
+ "model.layers.22.mlp.down_proj.weight": "model-00009-of-00044.safetensors",
325
+ "model.layers.22.mlp.gate_proj.weight": "model-00009-of-00044.safetensors",
326
+ "model.layers.22.mlp.up_proj.weight": "model-00009-of-00044.safetensors",
327
+ "model.layers.22.post_attention_layernorm.weight": "model-00009-of-00044.safetensors",
328
+ "model.layers.22.self_attn.k_proj.weight": "model-00009-of-00044.safetensors",
329
+ "model.layers.22.self_attn.o_proj.weight": "model-00009-of-00044.safetensors",
330
+ "model.layers.22.self_attn.q_proj.weight": "model-00009-of-00044.safetensors",
331
+ "model.layers.22.self_attn.v_proj.weight": "model-00009-of-00044.safetensors",
332
+ "model.layers.23.input_layernorm.weight": "model-00009-of-00044.safetensors",
333
+ "model.layers.23.mlp.down_proj.weight": "model-00009-of-00044.safetensors",
334
+ "model.layers.23.mlp.gate_proj.weight": "model-00009-of-00044.safetensors",
335
+ "model.layers.23.mlp.up_proj.weight": "model-00009-of-00044.safetensors",
336
+ "model.layers.23.post_attention_layernorm.weight": "model-00009-of-00044.safetensors",
337
+ "model.layers.23.self_attn.k_proj.weight": "model-00009-of-00044.safetensors",
338
+ "model.layers.23.self_attn.o_proj.weight": "model-00009-of-00044.safetensors",
339
+ "model.layers.23.self_attn.q_proj.weight": "model-00009-of-00044.safetensors",
340
+ "model.layers.23.self_attn.v_proj.weight": "model-00009-of-00044.safetensors",
341
+ "model.layers.24.input_layernorm.weight": "model-00010-of-00044.safetensors",
342
+ "model.layers.24.mlp.down_proj.weight": "model-00010-of-00044.safetensors",
343
+ "model.layers.24.mlp.gate_proj.weight": "model-00010-of-00044.safetensors",
344
+ "model.layers.24.mlp.up_proj.weight": "model-00010-of-00044.safetensors",
345
+ "model.layers.24.post_attention_layernorm.weight": "model-00010-of-00044.safetensors",
346
+ "model.layers.24.self_attn.k_proj.weight": "model-00010-of-00044.safetensors",
347
+ "model.layers.24.self_attn.o_proj.weight": "model-00010-of-00044.safetensors",
348
+ "model.layers.24.self_attn.q_proj.weight": "model-00010-of-00044.safetensors",
349
+ "model.layers.24.self_attn.v_proj.weight": "model-00010-of-00044.safetensors",
350
+ "model.layers.25.input_layernorm.weight": "model-00010-of-00044.safetensors",
351
+ "model.layers.25.mlp.down_proj.weight": "model-00010-of-00044.safetensors",
352
+ "model.layers.25.mlp.gate_proj.weight": "model-00010-of-00044.safetensors",
353
+ "model.layers.25.mlp.up_proj.weight": "model-00010-of-00044.safetensors",
354
+ "model.layers.25.post_attention_layernorm.weight": "model-00010-of-00044.safetensors",
355
+ "model.layers.25.self_attn.k_proj.weight": "model-00010-of-00044.safetensors",
356
+ "model.layers.25.self_attn.o_proj.weight": "model-00010-of-00044.safetensors",
357
+ "model.layers.25.self_attn.q_proj.weight": "model-00010-of-00044.safetensors",
358
+ "model.layers.25.self_attn.v_proj.weight": "model-00010-of-00044.safetensors",
359
+ "model.layers.26.input_layernorm.weight": "model-00011-of-00044.safetensors",
360
+ "model.layers.26.mlp.down_proj.weight": "model-00011-of-00044.safetensors",
361
+ "model.layers.26.mlp.gate_proj.weight": "model-00010-of-00044.safetensors",
362
+ "model.layers.26.mlp.up_proj.weight": "model-00010-of-00044.safetensors",
363
+ "model.layers.26.post_attention_layernorm.weight": "model-00011-of-00044.safetensors",
364
+ "model.layers.26.self_attn.k_proj.weight": "model-00010-of-00044.safetensors",
365
+ "model.layers.26.self_attn.o_proj.weight": "model-00010-of-00044.safetensors",
366
+ "model.layers.26.self_attn.q_proj.weight": "model-00010-of-00044.safetensors",
367
+ "model.layers.26.self_attn.v_proj.weight": "model-00010-of-00044.safetensors",
368
+ "model.layers.27.input_layernorm.weight": "model-00011-of-00044.safetensors",
369
+ "model.layers.27.mlp.down_proj.weight": "model-00011-of-00044.safetensors",
370
+ "model.layers.27.mlp.gate_proj.weight": "model-00011-of-00044.safetensors",
371
+ "model.layers.27.mlp.up_proj.weight": "model-00011-of-00044.safetensors",
372
+ "model.layers.27.post_attention_layernorm.weight": "model-00011-of-00044.safetensors",
373
+ "model.layers.27.self_attn.k_proj.weight": "model-00011-of-00044.safetensors",
374
+ "model.layers.27.self_attn.o_proj.weight": "model-00011-of-00044.safetensors",
375
+ "model.layers.27.self_attn.q_proj.weight": "model-00011-of-00044.safetensors",
376
+ "model.layers.27.self_attn.v_proj.weight": "model-00011-of-00044.safetensors",
377
+ "model.layers.28.input_layernorm.weight": "model-00011-of-00044.safetensors",
378
+ "model.layers.28.mlp.down_proj.weight": "model-00011-of-00044.safetensors",
379
+ "model.layers.28.mlp.gate_proj.weight": "model-00011-of-00044.safetensors",
380
+ "model.layers.28.mlp.up_proj.weight": "model-00011-of-00044.safetensors",
381
+ "model.layers.28.post_attention_layernorm.weight": "model-00011-of-00044.safetensors",
382
+ "model.layers.28.self_attn.k_proj.weight": "model-00011-of-00044.safetensors",
383
+ "model.layers.28.self_attn.o_proj.weight": "model-00011-of-00044.safetensors",
384
+ "model.layers.28.self_attn.q_proj.weight": "model-00011-of-00044.safetensors",
385
+ "model.layers.28.self_attn.v_proj.weight": "model-00011-of-00044.safetensors",
386
+ "model.layers.29.input_layernorm.weight": "model-00012-of-00044.safetensors",
387
+ "model.layers.29.mlp.down_proj.weight": "model-00012-of-00044.safetensors",
388
+ "model.layers.29.mlp.gate_proj.weight": "model-00011-of-00044.safetensors",
389
+ "model.layers.29.mlp.up_proj.weight": "model-00012-of-00044.safetensors",
390
+ "model.layers.29.post_attention_layernorm.weight": "model-00012-of-00044.safetensors",
391
+ "model.layers.29.self_attn.k_proj.weight": "model-00011-of-00044.safetensors",
392
+ "model.layers.29.self_attn.o_proj.weight": "model-00011-of-00044.safetensors",
393
+ "model.layers.29.self_attn.q_proj.weight": "model-00011-of-00044.safetensors",
394
+ "model.layers.29.self_attn.v_proj.weight": "model-00011-of-00044.safetensors",
395
+ "model.layers.3.input_layernorm.weight": "model-00002-of-00044.safetensors",
396
+ "model.layers.3.mlp.down_proj.weight": "model-00002-of-00044.safetensors",
397
+ "model.layers.3.mlp.gate_proj.weight": "model-00002-of-00044.safetensors",
398
+ "model.layers.3.mlp.up_proj.weight": "model-00002-of-00044.safetensors",
399
+ "model.layers.3.post_attention_layernorm.weight": "model-00002-of-00044.safetensors",
400
+ "model.layers.3.self_attn.k_proj.weight": "model-00002-of-00044.safetensors",
401
+ "model.layers.3.self_attn.o_proj.weight": "model-00002-of-00044.safetensors",
402
+ "model.layers.3.self_attn.q_proj.weight": "model-00002-of-00044.safetensors",
403
+ "model.layers.3.self_attn.v_proj.weight": "model-00002-of-00044.safetensors",
404
+ "model.layers.30.input_layernorm.weight": "model-00012-of-00044.safetensors",
405
+ "model.layers.30.mlp.down_proj.weight": "model-00012-of-00044.safetensors",
406
+ "model.layers.30.mlp.gate_proj.weight": "model-00012-of-00044.safetensors",
407
+ "model.layers.30.mlp.up_proj.weight": "model-00012-of-00044.safetensors",
408
+ "model.layers.30.post_attention_layernorm.weight": "model-00012-of-00044.safetensors",
409
+ "model.layers.30.self_attn.k_proj.weight": "model-00012-of-00044.safetensors",
410
+ "model.layers.30.self_attn.o_proj.weight": "model-00012-of-00044.safetensors",
411
+ "model.layers.30.self_attn.q_proj.weight": "model-00012-of-00044.safetensors",
412
+ "model.layers.30.self_attn.v_proj.weight": "model-00012-of-00044.safetensors",
413
+ "model.layers.31.input_layernorm.weight": "model-00012-of-00044.safetensors",
414
+ "model.layers.31.mlp.down_proj.weight": "model-00012-of-00044.safetensors",
415
+ "model.layers.31.mlp.gate_proj.weight": "model-00012-of-00044.safetensors",
416
+ "model.layers.31.mlp.up_proj.weight": "model-00012-of-00044.safetensors",
417
+ "model.layers.31.post_attention_layernorm.weight": "model-00012-of-00044.safetensors",
418
+ "model.layers.31.self_attn.k_proj.weight": "model-00012-of-00044.safetensors",
419
+ "model.layers.31.self_attn.o_proj.weight": "model-00012-of-00044.safetensors",
420
+ "model.layers.31.self_attn.q_proj.weight": "model-00012-of-00044.safetensors",
421
+ "model.layers.31.self_attn.v_proj.weight": "model-00012-of-00044.safetensors",
422
+ "model.layers.32.input_layernorm.weight": "model-00013-of-00044.safetensors",
423
+ "model.layers.32.mlp.down_proj.weight": "model-00013-of-00044.safetensors",
424
+ "model.layers.32.mlp.gate_proj.weight": "model-00013-of-00044.safetensors",
425
+ "model.layers.32.mlp.up_proj.weight": "model-00013-of-00044.safetensors",
426
+ "model.layers.32.post_attention_layernorm.weight": "model-00013-of-00044.safetensors",
427
+ "model.layers.32.self_attn.k_proj.weight": "model-00012-of-00044.safetensors",
428
+ "model.layers.32.self_attn.o_proj.weight": "model-00012-of-00044.safetensors",
429
+ "model.layers.32.self_attn.q_proj.weight": "model-00012-of-00044.safetensors",
430
+ "model.layers.32.self_attn.v_proj.weight": "model-00012-of-00044.safetensors",
431
+ "model.layers.33.input_layernorm.weight": "model-00013-of-00044.safetensors",
432
+ "model.layers.33.mlp.down_proj.weight": "model-00013-of-00044.safetensors",
433
+ "model.layers.33.mlp.gate_proj.weight": "model-00013-of-00044.safetensors",
434
+ "model.layers.33.mlp.up_proj.weight": "model-00013-of-00044.safetensors",
435
+ "model.layers.33.post_attention_layernorm.weight": "model-00013-of-00044.safetensors",
436
+ "model.layers.33.self_attn.k_proj.weight": "model-00013-of-00044.safetensors",
437
+ "model.layers.33.self_attn.o_proj.weight": "model-00013-of-00044.safetensors",
438
+ "model.layers.33.self_attn.q_proj.weight": "model-00013-of-00044.safetensors",
439
+ "model.layers.33.self_attn.v_proj.weight": "model-00013-of-00044.safetensors",
440
+ "model.layers.34.input_layernorm.weight": "model-00013-of-00044.safetensors",
441
+ "model.layers.34.mlp.down_proj.weight": "model-00013-of-00044.safetensors",
442
+ "model.layers.34.mlp.gate_proj.weight": "model-00013-of-00044.safetensors",
443
+ "model.layers.34.mlp.up_proj.weight": "model-00013-of-00044.safetensors",
444
+ "model.layers.34.post_attention_layernorm.weight": "model-00013-of-00044.safetensors",
445
+ "model.layers.34.self_attn.k_proj.weight": "model-00013-of-00044.safetensors",
446
+ "model.layers.34.self_attn.o_proj.weight": "model-00013-of-00044.safetensors",
447
+ "model.layers.34.self_attn.q_proj.weight": "model-00013-of-00044.safetensors",
448
+ "model.layers.34.self_attn.v_proj.weight": "model-00013-of-00044.safetensors",
449
+ "model.layers.35.input_layernorm.weight": "model-00014-of-00044.safetensors",
450
+ "model.layers.35.mlp.down_proj.weight": "model-00014-of-00044.safetensors",
451
+ "model.layers.35.mlp.gate_proj.weight": "model-00014-of-00044.safetensors",
452
+ "model.layers.35.mlp.up_proj.weight": "model-00014-of-00044.safetensors",
453
+ "model.layers.35.post_attention_layernorm.weight": "model-00014-of-00044.safetensors",
454
+ "model.layers.35.self_attn.k_proj.weight": "model-00013-of-00044.safetensors",
455
+ "model.layers.35.self_attn.o_proj.weight": "model-00014-of-00044.safetensors",
456
+ "model.layers.35.self_attn.q_proj.weight": "model-00013-of-00044.safetensors",
457
+ "model.layers.35.self_attn.v_proj.weight": "model-00013-of-00044.safetensors",
458
+ "model.layers.36.input_layernorm.weight": "model-00014-of-00044.safetensors",
459
+ "model.layers.36.mlp.down_proj.weight": "model-00014-of-00044.safetensors",
460
+ "model.layers.36.mlp.gate_proj.weight": "model-00014-of-00044.safetensors",
461
+ "model.layers.36.mlp.up_proj.weight": "model-00014-of-00044.safetensors",
462
+ "model.layers.36.post_attention_layernorm.weight": "model-00014-of-00044.safetensors",
463
+ "model.layers.36.self_attn.k_proj.weight": "model-00014-of-00044.safetensors",
464
+ "model.layers.36.self_attn.o_proj.weight": "model-00014-of-00044.safetensors",
465
+ "model.layers.36.self_attn.q_proj.weight": "model-00014-of-00044.safetensors",
466
+ "model.layers.36.self_attn.v_proj.weight": "model-00014-of-00044.safetensors",
467
+ "model.layers.37.input_layernorm.weight": "model-00014-of-00044.safetensors",
468
+ "model.layers.37.mlp.down_proj.weight": "model-00014-of-00044.safetensors",
469
+ "model.layers.37.mlp.gate_proj.weight": "model-00014-of-00044.safetensors",
470
+ "model.layers.37.mlp.up_proj.weight": "model-00014-of-00044.safetensors",
471
+ "model.layers.37.post_attention_layernorm.weight": "model-00014-of-00044.safetensors",
472
+ "model.layers.37.self_attn.k_proj.weight": "model-00014-of-00044.safetensors",
473
+ "model.layers.37.self_attn.o_proj.weight": "model-00014-of-00044.safetensors",
474
+ "model.layers.37.self_attn.q_proj.weight": "model-00014-of-00044.safetensors",
475
+ "model.layers.37.self_attn.v_proj.weight": "model-00014-of-00044.safetensors",
476
+ "model.layers.38.input_layernorm.weight": "model-00015-of-00044.safetensors",
477
+ "model.layers.38.mlp.down_proj.weight": "model-00015-of-00044.safetensors",
478
+ "model.layers.38.mlp.gate_proj.weight": "model-00015-of-00044.safetensors",
479
+ "model.layers.38.mlp.up_proj.weight": "model-00015-of-00044.safetensors",
480
+ "model.layers.38.post_attention_layernorm.weight": "model-00015-of-00044.safetensors",
481
+ "model.layers.38.self_attn.k_proj.weight": "model-00015-of-00044.safetensors",
482
+ "model.layers.38.self_attn.o_proj.weight": "model-00015-of-00044.safetensors",
483
+ "model.layers.38.self_attn.q_proj.weight": "model-00015-of-00044.safetensors",
484
+ "model.layers.38.self_attn.v_proj.weight": "model-00015-of-00044.safetensors",
485
+ "model.layers.39.input_layernorm.weight": "model-00015-of-00044.safetensors",
486
+ "model.layers.39.mlp.down_proj.weight": "model-00015-of-00044.safetensors",
487
+ "model.layers.39.mlp.gate_proj.weight": "model-00015-of-00044.safetensors",
488
+ "model.layers.39.mlp.up_proj.weight": "model-00015-of-00044.safetensors",
489
+ "model.layers.39.post_attention_layernorm.weight": "model-00015-of-00044.safetensors",
490
+ "model.layers.39.self_attn.k_proj.weight": "model-00015-of-00044.safetensors",
491
+ "model.layers.39.self_attn.o_proj.weight": "model-00015-of-00044.safetensors",
492
+ "model.layers.39.self_attn.q_proj.weight": "model-00015-of-00044.safetensors",
493
+ "model.layers.39.self_attn.v_proj.weight": "model-00015-of-00044.safetensors",
494
+ "model.layers.4.input_layernorm.weight": "model-00003-of-00044.safetensors",
495
+ "model.layers.4.mlp.down_proj.weight": "model-00003-of-00044.safetensors",
496
+ "model.layers.4.mlp.gate_proj.weight": "model-00003-of-00044.safetensors",
497
+ "model.layers.4.mlp.up_proj.weight": "model-00003-of-00044.safetensors",
498
+ "model.layers.4.post_attention_layernorm.weight": "model-00003-of-00044.safetensors",
499
+ "model.layers.4.self_attn.k_proj.weight": "model-00002-of-00044.safetensors",
500
+ "model.layers.4.self_attn.o_proj.weight": "model-00002-of-00044.safetensors",
501
+ "model.layers.4.self_attn.q_proj.weight": "model-00002-of-00044.safetensors",
502
+ "model.layers.4.self_attn.v_proj.weight": "model-00002-of-00044.safetensors",
503
+ "model.layers.40.input_layernorm.weight": "model-00016-of-00044.safetensors",
504
+ "model.layers.40.mlp.down_proj.weight": "model-00016-of-00044.safetensors",
505
+ "model.layers.40.mlp.gate_proj.weight": "model-00015-of-00044.safetensors",
506
+ "model.layers.40.mlp.up_proj.weight": "model-00015-of-00044.safetensors",
507
+ "model.layers.40.post_attention_layernorm.weight": "model-00016-of-00044.safetensors",
508
+ "model.layers.40.self_attn.k_proj.weight": "model-00015-of-00044.safetensors",
509
+ "model.layers.40.self_attn.o_proj.weight": "model-00015-of-00044.safetensors",
510
+ "model.layers.40.self_attn.q_proj.weight": "model-00015-of-00044.safetensors",
511
+ "model.layers.40.self_attn.v_proj.weight": "model-00015-of-00044.safetensors",
512
+ "model.layers.41.input_layernorm.weight": "model-00016-of-00044.safetensors",
513
+ "model.layers.41.mlp.down_proj.weight": "model-00016-of-00044.safetensors",
514
+ "model.layers.41.mlp.gate_proj.weight": "model-00016-of-00044.safetensors",
515
+ "model.layers.41.mlp.up_proj.weight": "model-00016-of-00044.safetensors",
516
+ "model.layers.41.post_attention_layernorm.weight": "model-00016-of-00044.safetensors",
517
+ "model.layers.41.self_attn.k_proj.weight": "model-00016-of-00044.safetensors",
518
+ "model.layers.41.self_attn.o_proj.weight": "model-00016-of-00044.safetensors",
519
+ "model.layers.41.self_attn.q_proj.weight": "model-00016-of-00044.safetensors",
520
+ "model.layers.41.self_attn.v_proj.weight": "model-00016-of-00044.safetensors",
521
+ "model.layers.42.input_layernorm.weight": "model-00016-of-00044.safetensors",
522
+ "model.layers.42.mlp.down_proj.weight": "model-00016-of-00044.safetensors",
523
+ "model.layers.42.mlp.gate_proj.weight": "model-00016-of-00044.safetensors",
524
+ "model.layers.42.mlp.up_proj.weight": "model-00016-of-00044.safetensors",
525
+ "model.layers.42.post_attention_layernorm.weight": "model-00016-of-00044.safetensors",
526
+ "model.layers.42.self_attn.k_proj.weight": "model-00016-of-00044.safetensors",
527
+ "model.layers.42.self_attn.o_proj.weight": "model-00016-of-00044.safetensors",
528
+ "model.layers.42.self_attn.q_proj.weight": "model-00016-of-00044.safetensors",
529
+ "model.layers.42.self_attn.v_proj.weight": "model-00016-of-00044.safetensors",
530
+ "model.layers.43.input_layernorm.weight": "model-00017-of-00044.safetensors",
531
+ "model.layers.43.mlp.down_proj.weight": "model-00017-of-00044.safetensors",
532
+ "model.layers.43.mlp.gate_proj.weight": "model-00016-of-00044.safetensors",
533
+ "model.layers.43.mlp.up_proj.weight": "model-00017-of-00044.safetensors",
534
+ "model.layers.43.post_attention_layernorm.weight": "model-00017-of-00044.safetensors",
535
+ "model.layers.43.self_attn.k_proj.weight": "model-00016-of-00044.safetensors",
536
+ "model.layers.43.self_attn.o_proj.weight": "model-00016-of-00044.safetensors",
537
+ "model.layers.43.self_attn.q_proj.weight": "model-00016-of-00044.safetensors",
538
+ "model.layers.43.self_attn.v_proj.weight": "model-00016-of-00044.safetensors",
539
+ "model.layers.44.input_layernorm.weight": "model-00017-of-00044.safetensors",
540
+ "model.layers.44.mlp.down_proj.weight": "model-00017-of-00044.safetensors",
541
+ "model.layers.44.mlp.gate_proj.weight": "model-00017-of-00044.safetensors",
542
+ "model.layers.44.mlp.up_proj.weight": "model-00017-of-00044.safetensors",
543
+ "model.layers.44.post_attention_layernorm.weight": "model-00017-of-00044.safetensors",
544
+ "model.layers.44.self_attn.k_proj.weight": "model-00017-of-00044.safetensors",
545
+ "model.layers.44.self_attn.o_proj.weight": "model-00017-of-00044.safetensors",
546
+ "model.layers.44.self_attn.q_proj.weight": "model-00017-of-00044.safetensors",
547
+ "model.layers.44.self_attn.v_proj.weight": "model-00017-of-00044.safetensors",
548
+ "model.layers.45.input_layernorm.weight": "model-00017-of-00044.safetensors",
549
+ "model.layers.45.mlp.down_proj.weight": "model-00017-of-00044.safetensors",
550
+ "model.layers.45.mlp.gate_proj.weight": "model-00017-of-00044.safetensors",
551
+ "model.layers.45.mlp.up_proj.weight": "model-00017-of-00044.safetensors",
552
+ "model.layers.45.post_attention_layernorm.weight": "model-00017-of-00044.safetensors",
553
+ "model.layers.45.self_attn.k_proj.weight": "model-00017-of-00044.safetensors",
554
+ "model.layers.45.self_attn.o_proj.weight": "model-00017-of-00044.safetensors",
555
+ "model.layers.45.self_attn.q_proj.weight": "model-00017-of-00044.safetensors",
556
+ "model.layers.45.self_attn.v_proj.weight": "model-00017-of-00044.safetensors",
557
+ "model.layers.46.input_layernorm.weight": "model-00018-of-00044.safetensors",
558
+ "model.layers.46.mlp.down_proj.weight": "model-00018-of-00044.safetensors",
559
+ "model.layers.46.mlp.gate_proj.weight": "model-00018-of-00044.safetensors",
560
+ "model.layers.46.mlp.up_proj.weight": "model-00018-of-00044.safetensors",
561
+ "model.layers.46.post_attention_layernorm.weight": "model-00018-of-00044.safetensors",
562
+ "model.layers.46.self_attn.k_proj.weight": "model-00017-of-00044.safetensors",
563
+ "model.layers.46.self_attn.o_proj.weight": "model-00017-of-00044.safetensors",
564
+ "model.layers.46.self_attn.q_proj.weight": "model-00017-of-00044.safetensors",
565
+ "model.layers.46.self_attn.v_proj.weight": "model-00017-of-00044.safetensors",
566
+ "model.layers.47.input_layernorm.weight": "model-00018-of-00044.safetensors",
567
+ "model.layers.47.mlp.down_proj.weight": "model-00018-of-00044.safetensors",
568
+ "model.layers.47.mlp.gate_proj.weight": "model-00018-of-00044.safetensors",
569
+ "model.layers.47.mlp.up_proj.weight": "model-00018-of-00044.safetensors",
570
+ "model.layers.47.post_attention_layernorm.weight": "model-00018-of-00044.safetensors",
571
+ "model.layers.47.self_attn.k_proj.weight": "model-00018-of-00044.safetensors",
572
+ "model.layers.47.self_attn.o_proj.weight": "model-00018-of-00044.safetensors",
573
+ "model.layers.47.self_attn.q_proj.weight": "model-00018-of-00044.safetensors",
574
+ "model.layers.47.self_attn.v_proj.weight": "model-00018-of-00044.safetensors",
575
+ "model.layers.48.input_layernorm.weight": "model-00018-of-00044.safetensors",
576
+ "model.layers.48.mlp.down_proj.weight": "model-00018-of-00044.safetensors",
577
+ "model.layers.48.mlp.gate_proj.weight": "model-00018-of-00044.safetensors",
578
+ "model.layers.48.mlp.up_proj.weight": "model-00018-of-00044.safetensors",
579
+ "model.layers.48.post_attention_layernorm.weight": "model-00018-of-00044.safetensors",
580
+ "model.layers.48.self_attn.k_proj.weight": "model-00018-of-00044.safetensors",
581
+ "model.layers.48.self_attn.o_proj.weight": "model-00018-of-00044.safetensors",
582
+ "model.layers.48.self_attn.q_proj.weight": "model-00018-of-00044.safetensors",
583
+ "model.layers.48.self_attn.v_proj.weight": "model-00018-of-00044.safetensors",
584
+ "model.layers.49.input_layernorm.weight": "model-00019-of-00044.safetensors",
585
+ "model.layers.49.mlp.down_proj.weight": "model-00019-of-00044.safetensors",
586
+ "model.layers.49.mlp.gate_proj.weight": "model-00019-of-00044.safetensors",
587
+ "model.layers.49.mlp.up_proj.weight": "model-00019-of-00044.safetensors",
588
+ "model.layers.49.post_attention_layernorm.weight": "model-00019-of-00044.safetensors",
589
+ "model.layers.49.self_attn.k_proj.weight": "model-00018-of-00044.safetensors",
590
+ "model.layers.49.self_attn.o_proj.weight": "model-00019-of-00044.safetensors",
591
+ "model.layers.49.self_attn.q_proj.weight": "model-00018-of-00044.safetensors",
592
+ "model.layers.49.self_attn.v_proj.weight": "model-00018-of-00044.safetensors",
593
+ "model.layers.5.input_layernorm.weight": "model-00003-of-00044.safetensors",
594
+ "model.layers.5.mlp.down_proj.weight": "model-00003-of-00044.safetensors",
595
+ "model.layers.5.mlp.gate_proj.weight": "model-00003-of-00044.safetensors",
596
+ "model.layers.5.mlp.up_proj.weight": "model-00003-of-00044.safetensors",
597
+ "model.layers.5.post_attention_layernorm.weight": "model-00003-of-00044.safetensors",
598
+ "model.layers.5.self_attn.k_proj.weight": "model-00003-of-00044.safetensors",
599
+ "model.layers.5.self_attn.o_proj.weight": "model-00003-of-00044.safetensors",
600
+ "model.layers.5.self_attn.q_proj.weight": "model-00003-of-00044.safetensors",
601
+ "model.layers.5.self_attn.v_proj.weight": "model-00003-of-00044.safetensors",
602
+ "model.layers.50.input_layernorm.weight": "model-00019-of-00044.safetensors",
603
+ "model.layers.50.mlp.down_proj.weight": "model-00019-of-00044.safetensors",
604
+ "model.layers.50.mlp.gate_proj.weight": "model-00019-of-00044.safetensors",
605
+ "model.layers.50.mlp.up_proj.weight": "model-00019-of-00044.safetensors",
606
+ "model.layers.50.post_attention_layernorm.weight": "model-00019-of-00044.safetensors",
607
+ "model.layers.50.self_attn.k_proj.weight": "model-00019-of-00044.safetensors",
608
+ "model.layers.50.self_attn.o_proj.weight": "model-00019-of-00044.safetensors",
609
+ "model.layers.50.self_attn.q_proj.weight": "model-00019-of-00044.safetensors",
610
+ "model.layers.50.self_attn.v_proj.weight": "model-00019-of-00044.safetensors",
611
+ "model.layers.51.input_layernorm.weight": "model-00019-of-00044.safetensors",
612
+ "model.layers.51.mlp.down_proj.weight": "model-00019-of-00044.safetensors",
613
+ "model.layers.51.mlp.gate_proj.weight": "model-00019-of-00044.safetensors",
614
+ "model.layers.51.mlp.up_proj.weight": "model-00019-of-00044.safetensors",
615
+ "model.layers.51.post_attention_layernorm.weight": "model-00019-of-00044.safetensors",
616
+ "model.layers.51.self_attn.k_proj.weight": "model-00019-of-00044.safetensors",
617
+ "model.layers.51.self_attn.o_proj.weight": "model-00019-of-00044.safetensors",
618
+ "model.layers.51.self_attn.q_proj.weight": "model-00019-of-00044.safetensors",
619
+ "model.layers.51.self_attn.v_proj.weight": "model-00019-of-00044.safetensors",
620
+ "model.layers.52.input_layernorm.weight": "model-00020-of-00044.safetensors",
621
+ "model.layers.52.mlp.down_proj.weight": "model-00020-of-00044.safetensors",
622
+ "model.layers.52.mlp.gate_proj.weight": "model-00020-of-00044.safetensors",
623
+ "model.layers.52.mlp.up_proj.weight": "model-00020-of-00044.safetensors",
624
+ "model.layers.52.post_attention_layernorm.weight": "model-00020-of-00044.safetensors",
625
+ "model.layers.52.self_attn.k_proj.weight": "model-00020-of-00044.safetensors",
626
+ "model.layers.52.self_attn.o_proj.weight": "model-00020-of-00044.safetensors",
627
+ "model.layers.52.self_attn.q_proj.weight": "model-00020-of-00044.safetensors",
628
+ "model.layers.52.self_attn.v_proj.weight": "model-00020-of-00044.safetensors",
629
+ "model.layers.53.input_layernorm.weight": "model-00020-of-00044.safetensors",
630
+ "model.layers.53.mlp.down_proj.weight": "model-00020-of-00044.safetensors",
631
+ "model.layers.53.mlp.gate_proj.weight": "model-00020-of-00044.safetensors",
632
+ "model.layers.53.mlp.up_proj.weight": "model-00020-of-00044.safetensors",
633
+ "model.layers.53.post_attention_layernorm.weight": "model-00020-of-00044.safetensors",
634
+ "model.layers.53.self_attn.k_proj.weight": "model-00020-of-00044.safetensors",
635
+ "model.layers.53.self_attn.o_proj.weight": "model-00020-of-00044.safetensors",
636
+ "model.layers.53.self_attn.q_proj.weight": "model-00020-of-00044.safetensors",
637
+ "model.layers.53.self_attn.v_proj.weight": "model-00020-of-00044.safetensors",
638
+ "model.layers.54.input_layernorm.weight": "model-00021-of-00044.safetensors",
639
+ "model.layers.54.mlp.down_proj.weight": "model-00021-of-00044.safetensors",
640
+ "model.layers.54.mlp.gate_proj.weight": "model-00020-of-00044.safetensors",
641
+ "model.layers.54.mlp.up_proj.weight": "model-00020-of-00044.safetensors",
642
+ "model.layers.54.post_attention_layernorm.weight": "model-00021-of-00044.safetensors",
643
+ "model.layers.54.self_attn.k_proj.weight": "model-00020-of-00044.safetensors",
644
+ "model.layers.54.self_attn.o_proj.weight": "model-00020-of-00044.safetensors",
645
+ "model.layers.54.self_attn.q_proj.weight": "model-00020-of-00044.safetensors",
646
+ "model.layers.54.self_attn.v_proj.weight": "model-00020-of-00044.safetensors",
647
+ "model.layers.55.input_layernorm.weight": "model-00021-of-00044.safetensors",
648
+ "model.layers.55.mlp.down_proj.weight": "model-00021-of-00044.safetensors",
649
+ "model.layers.55.mlp.gate_proj.weight": "model-00021-of-00044.safetensors",
650
+ "model.layers.55.mlp.up_proj.weight": "model-00021-of-00044.safetensors",
651
+ "model.layers.55.post_attention_layernorm.weight": "model-00021-of-00044.safetensors",
652
+ "model.layers.55.self_attn.k_proj.weight": "model-00021-of-00044.safetensors",
653
+ "model.layers.55.self_attn.o_proj.weight": "model-00021-of-00044.safetensors",
654
+ "model.layers.55.self_attn.q_proj.weight": "model-00021-of-00044.safetensors",
655
+ "model.layers.55.self_attn.v_proj.weight": "model-00021-of-00044.safetensors",
656
+ "model.layers.56.input_layernorm.weight": "model-00021-of-00044.safetensors",
657
+ "model.layers.56.mlp.down_proj.weight": "model-00021-of-00044.safetensors",
658
+ "model.layers.56.mlp.gate_proj.weight": "model-00021-of-00044.safetensors",
659
+ "model.layers.56.mlp.up_proj.weight": "model-00021-of-00044.safetensors",
660
+ "model.layers.56.post_attention_layernorm.weight": "model-00021-of-00044.safetensors",
661
+ "model.layers.56.self_attn.k_proj.weight": "model-00021-of-00044.safetensors",
662
+ "model.layers.56.self_attn.o_proj.weight": "model-00021-of-00044.safetensors",
663
+ "model.layers.56.self_attn.q_proj.weight": "model-00021-of-00044.safetensors",
664
+ "model.layers.56.self_attn.v_proj.weight": "model-00021-of-00044.safetensors",
665
+ "model.layers.57.input_layernorm.weight": "model-00022-of-00044.safetensors",
666
+ "model.layers.57.mlp.down_proj.weight": "model-00022-of-00044.safetensors",
667
+ "model.layers.57.mlp.gate_proj.weight": "model-00021-of-00044.safetensors",
668
+ "model.layers.57.mlp.up_proj.weight": "model-00022-of-00044.safetensors",
669
+ "model.layers.57.post_attention_layernorm.weight": "model-00022-of-00044.safetensors",
670
+ "model.layers.57.self_attn.k_proj.weight": "model-00021-of-00044.safetensors",
671
+ "model.layers.57.self_attn.o_proj.weight": "model-00021-of-00044.safetensors",
672
+ "model.layers.57.self_attn.q_proj.weight": "model-00021-of-00044.safetensors",
673
+ "model.layers.57.self_attn.v_proj.weight": "model-00021-of-00044.safetensors",
674
+ "model.layers.58.input_layernorm.weight": "model-00022-of-00044.safetensors",
675
+ "model.layers.58.mlp.down_proj.weight": "model-00022-of-00044.safetensors",
676
+ "model.layers.58.mlp.gate_proj.weight": "model-00022-of-00044.safetensors",
677
+ "model.layers.58.mlp.up_proj.weight": "model-00022-of-00044.safetensors",
678
+ "model.layers.58.post_attention_layernorm.weight": "model-00022-of-00044.safetensors",
679
+ "model.layers.58.self_attn.k_proj.weight": "model-00022-of-00044.safetensors",
680
+ "model.layers.58.self_attn.o_proj.weight": "model-00022-of-00044.safetensors",
681
+ "model.layers.58.self_attn.q_proj.weight": "model-00022-of-00044.safetensors",
682
+ "model.layers.58.self_attn.v_proj.weight": "model-00022-of-00044.safetensors",
683
+ "model.layers.59.input_layernorm.weight": "model-00022-of-00044.safetensors",
684
+ "model.layers.59.mlp.down_proj.weight": "model-00022-of-00044.safetensors",
685
+ "model.layers.59.mlp.gate_proj.weight": "model-00022-of-00044.safetensors",
686
+ "model.layers.59.mlp.up_proj.weight": "model-00022-of-00044.safetensors",
687
+ "model.layers.59.post_attention_layernorm.weight": "model-00022-of-00044.safetensors",
688
+ "model.layers.59.self_attn.k_proj.weight": "model-00022-of-00044.safetensors",
689
+ "model.layers.59.self_attn.o_proj.weight": "model-00022-of-00044.safetensors",
690
+ "model.layers.59.self_attn.q_proj.weight": "model-00022-of-00044.safetensors",
691
+ "model.layers.59.self_attn.v_proj.weight": "model-00022-of-00044.safetensors",
692
+ "model.layers.6.input_layernorm.weight": "model-00003-of-00044.safetensors",
693
+ "model.layers.6.mlp.down_proj.weight": "model-00003-of-00044.safetensors",
694
+ "model.layers.6.mlp.gate_proj.weight": "model-00003-of-00044.safetensors",
695
+ "model.layers.6.mlp.up_proj.weight": "model-00003-of-00044.safetensors",
696
+ "model.layers.6.post_attention_layernorm.weight": "model-00003-of-00044.safetensors",
697
+ "model.layers.6.self_attn.k_proj.weight": "model-00003-of-00044.safetensors",
698
+ "model.layers.6.self_attn.o_proj.weight": "model-00003-of-00044.safetensors",
699
+ "model.layers.6.self_attn.q_proj.weight": "model-00003-of-00044.safetensors",
700
+ "model.layers.6.self_attn.v_proj.weight": "model-00003-of-00044.safetensors",
701
+ "model.layers.60.input_layernorm.weight": "model-00023-of-00044.safetensors",
702
+ "model.layers.60.mlp.down_proj.weight": "model-00023-of-00044.safetensors",
703
+ "model.layers.60.mlp.gate_proj.weight": "model-00023-of-00044.safetensors",
704
+ "model.layers.60.mlp.up_proj.weight": "model-00023-of-00044.safetensors",
705
+ "model.layers.60.post_attention_layernorm.weight": "model-00023-of-00044.safetensors",
706
+ "model.layers.60.self_attn.k_proj.weight": "model-00022-of-00044.safetensors",
707
+ "model.layers.60.self_attn.o_proj.weight": "model-00022-of-00044.safetensors",
708
+ "model.layers.60.self_attn.q_proj.weight": "model-00022-of-00044.safetensors",
709
+ "model.layers.60.self_attn.v_proj.weight": "model-00022-of-00044.safetensors",
710
+ "model.layers.61.input_layernorm.weight": "model-00023-of-00044.safetensors",
711
+ "model.layers.61.mlp.down_proj.weight": "model-00023-of-00044.safetensors",
712
+ "model.layers.61.mlp.gate_proj.weight": "model-00023-of-00044.safetensors",
713
+ "model.layers.61.mlp.up_proj.weight": "model-00023-of-00044.safetensors",
714
+ "model.layers.61.post_attention_layernorm.weight": "model-00023-of-00044.safetensors",
715
+ "model.layers.61.self_attn.k_proj.weight": "model-00023-of-00044.safetensors",
716
+ "model.layers.61.self_attn.o_proj.weight": "model-00023-of-00044.safetensors",
717
+ "model.layers.61.self_attn.q_proj.weight": "model-00023-of-00044.safetensors",
718
+ "model.layers.61.self_attn.v_proj.weight": "model-00023-of-00044.safetensors",
719
+ "model.layers.62.input_layernorm.weight": "model-00023-of-00044.safetensors",
720
+ "model.layers.62.mlp.down_proj.weight": "model-00023-of-00044.safetensors",
721
+ "model.layers.62.mlp.gate_proj.weight": "model-00023-of-00044.safetensors",
722
+ "model.layers.62.mlp.up_proj.weight": "model-00023-of-00044.safetensors",
723
+ "model.layers.62.post_attention_layernorm.weight": "model-00023-of-00044.safetensors",
724
+ "model.layers.62.self_attn.k_proj.weight": "model-00023-of-00044.safetensors",
725
+ "model.layers.62.self_attn.o_proj.weight": "model-00023-of-00044.safetensors",
726
+ "model.layers.62.self_attn.q_proj.weight": "model-00023-of-00044.safetensors",
727
+ "model.layers.62.self_attn.v_proj.weight": "model-00023-of-00044.safetensors",
728
+ "model.layers.63.input_layernorm.weight": "model-00024-of-00044.safetensors",
729
+ "model.layers.63.mlp.down_proj.weight": "model-00024-of-00044.safetensors",
730
+ "model.layers.63.mlp.gate_proj.weight": "model-00024-of-00044.safetensors",
731
+ "model.layers.63.mlp.up_proj.weight": "model-00024-of-00044.safetensors",
732
+ "model.layers.63.post_attention_layernorm.weight": "model-00024-of-00044.safetensors",
733
+ "model.layers.63.self_attn.k_proj.weight": "model-00023-of-00044.safetensors",
734
+ "model.layers.63.self_attn.o_proj.weight": "model-00024-of-00044.safetensors",
735
+ "model.layers.63.self_attn.q_proj.weight": "model-00023-of-00044.safetensors",
736
+ "model.layers.63.self_attn.v_proj.weight": "model-00023-of-00044.safetensors",
737
+ "model.layers.64.input_layernorm.weight": "model-00024-of-00044.safetensors",
738
+ "model.layers.64.mlp.down_proj.weight": "model-00024-of-00044.safetensors",
739
+ "model.layers.64.mlp.gate_proj.weight": "model-00024-of-00044.safetensors",
740
+ "model.layers.64.mlp.up_proj.weight": "model-00024-of-00044.safetensors",
741
+ "model.layers.64.post_attention_layernorm.weight": "model-00024-of-00044.safetensors",
742
+ "model.layers.64.self_attn.k_proj.weight": "model-00024-of-00044.safetensors",
743
+ "model.layers.64.self_attn.o_proj.weight": "model-00024-of-00044.safetensors",
744
+ "model.layers.64.self_attn.q_proj.weight": "model-00024-of-00044.safetensors",
745
+ "model.layers.64.self_attn.v_proj.weight": "model-00024-of-00044.safetensors",
746
+ "model.layers.65.input_layernorm.weight": "model-00024-of-00044.safetensors",
747
+ "model.layers.65.mlp.down_proj.weight": "model-00024-of-00044.safetensors",
748
+ "model.layers.65.mlp.gate_proj.weight": "model-00024-of-00044.safetensors",
749
+ "model.layers.65.mlp.up_proj.weight": "model-00024-of-00044.safetensors",
750
+ "model.layers.65.post_attention_layernorm.weight": "model-00024-of-00044.safetensors",
751
+ "model.layers.65.self_attn.k_proj.weight": "model-00024-of-00044.safetensors",
752
+ "model.layers.65.self_attn.o_proj.weight": "model-00024-of-00044.safetensors",
753
+ "model.layers.65.self_attn.q_proj.weight": "model-00024-of-00044.safetensors",
754
+ "model.layers.65.self_attn.v_proj.weight": "model-00024-of-00044.safetensors",
755
+ "model.layers.66.input_layernorm.weight": "model-00025-of-00044.safetensors",
756
+ "model.layers.66.mlp.down_proj.weight": "model-00025-of-00044.safetensors",
757
+ "model.layers.66.mlp.gate_proj.weight": "model-00025-of-00044.safetensors",
758
+ "model.layers.66.mlp.up_proj.weight": "model-00025-of-00044.safetensors",
759
+ "model.layers.66.post_attention_layernorm.weight": "model-00025-of-00044.safetensors",
760
+ "model.layers.66.self_attn.k_proj.weight": "model-00025-of-00044.safetensors",
761
+ "model.layers.66.self_attn.o_proj.weight": "model-00025-of-00044.safetensors",
762
+ "model.layers.66.self_attn.q_proj.weight": "model-00025-of-00044.safetensors",
763
+ "model.layers.66.self_attn.v_proj.weight": "model-00025-of-00044.safetensors",
764
+ "model.layers.67.input_layernorm.weight": "model-00025-of-00044.safetensors",
765
+ "model.layers.67.mlp.down_proj.weight": "model-00025-of-00044.safetensors",
766
+ "model.layers.67.mlp.gate_proj.weight": "model-00025-of-00044.safetensors",
767
+ "model.layers.67.mlp.up_proj.weight": "model-00025-of-00044.safetensors",
768
+ "model.layers.67.post_attention_layernorm.weight": "model-00025-of-00044.safetensors",
769
+ "model.layers.67.self_attn.k_proj.weight": "model-00025-of-00044.safetensors",
770
+ "model.layers.67.self_attn.o_proj.weight": "model-00025-of-00044.safetensors",
771
+ "model.layers.67.self_attn.q_proj.weight": "model-00025-of-00044.safetensors",
772
+ "model.layers.67.self_attn.v_proj.weight": "model-00025-of-00044.safetensors",
773
+ "model.layers.68.input_layernorm.weight": "model-00026-of-00044.safetensors",
774
+ "model.layers.68.mlp.down_proj.weight": "model-00026-of-00044.safetensors",
775
+ "model.layers.68.mlp.gate_proj.weight": "model-00025-of-00044.safetensors",
776
+ "model.layers.68.mlp.up_proj.weight": "model-00025-of-00044.safetensors",
777
+ "model.layers.68.post_attention_layernorm.weight": "model-00026-of-00044.safetensors",
778
+ "model.layers.68.self_attn.k_proj.weight": "model-00025-of-00044.safetensors",
779
+ "model.layers.68.self_attn.o_proj.weight": "model-00025-of-00044.safetensors",
780
+ "model.layers.68.self_attn.q_proj.weight": "model-00025-of-00044.safetensors",
781
+ "model.layers.68.self_attn.v_proj.weight": "model-00025-of-00044.safetensors",
782
+ "model.layers.69.input_layernorm.weight": "model-00026-of-00044.safetensors",
783
+ "model.layers.69.mlp.down_proj.weight": "model-00026-of-00044.safetensors",
784
+ "model.layers.69.mlp.gate_proj.weight": "model-00026-of-00044.safetensors",
785
+ "model.layers.69.mlp.up_proj.weight": "model-00026-of-00044.safetensors",
786
+ "model.layers.69.post_attention_layernorm.weight": "model-00026-of-00044.safetensors",
787
+ "model.layers.69.self_attn.k_proj.weight": "model-00026-of-00044.safetensors",
788
+ "model.layers.69.self_attn.o_proj.weight": "model-00026-of-00044.safetensors",
789
+ "model.layers.69.self_attn.q_proj.weight": "model-00026-of-00044.safetensors",
790
+ "model.layers.69.self_attn.v_proj.weight": "model-00026-of-00044.safetensors",
791
+ "model.layers.7.input_layernorm.weight": "model-00004-of-00044.safetensors",
792
+ "model.layers.7.mlp.down_proj.weight": "model-00004-of-00044.safetensors",
793
+ "model.layers.7.mlp.gate_proj.weight": "model-00004-of-00044.safetensors",
794
+ "model.layers.7.mlp.up_proj.weight": "model-00004-of-00044.safetensors",
795
+ "model.layers.7.post_attention_layernorm.weight": "model-00004-of-00044.safetensors",
796
+ "model.layers.7.self_attn.k_proj.weight": "model-00003-of-00044.safetensors",
797
+ "model.layers.7.self_attn.o_proj.weight": "model-00004-of-00044.safetensors",
798
+ "model.layers.7.self_attn.q_proj.weight": "model-00003-of-00044.safetensors",
799
+ "model.layers.7.self_attn.v_proj.weight": "model-00003-of-00044.safetensors",
800
+ "model.layers.70.input_layernorm.weight": "model-00026-of-00044.safetensors",
801
+ "model.layers.70.mlp.down_proj.weight": "model-00026-of-00044.safetensors",
802
+ "model.layers.70.mlp.gate_proj.weight": "model-00026-of-00044.safetensors",
803
+ "model.layers.70.mlp.up_proj.weight": "model-00026-of-00044.safetensors",
804
+ "model.layers.70.post_attention_layernorm.weight": "model-00026-of-00044.safetensors",
805
+ "model.layers.70.self_attn.k_proj.weight": "model-00026-of-00044.safetensors",
806
+ "model.layers.70.self_attn.o_proj.weight": "model-00026-of-00044.safetensors",
807
+ "model.layers.70.self_attn.q_proj.weight": "model-00026-of-00044.safetensors",
808
+ "model.layers.70.self_attn.v_proj.weight": "model-00026-of-00044.safetensors",
809
+ "model.layers.71.input_layernorm.weight": "model-00027-of-00044.safetensors",
810
+ "model.layers.71.mlp.down_proj.weight": "model-00027-of-00044.safetensors",
811
+ "model.layers.71.mlp.gate_proj.weight": "model-00026-of-00044.safetensors",
812
+ "model.layers.71.mlp.up_proj.weight": "model-00027-of-00044.safetensors",
813
+ "model.layers.71.post_attention_layernorm.weight": "model-00027-of-00044.safetensors",
814
+ "model.layers.71.self_attn.k_proj.weight": "model-00026-of-00044.safetensors",
815
+ "model.layers.71.self_attn.o_proj.weight": "model-00026-of-00044.safetensors",
816
+ "model.layers.71.self_attn.q_proj.weight": "model-00026-of-00044.safetensors",
817
+ "model.layers.71.self_attn.v_proj.weight": "model-00026-of-00044.safetensors",
818
+ "model.layers.72.input_layernorm.weight": "model-00027-of-00044.safetensors",
819
+ "model.layers.72.mlp.down_proj.weight": "model-00027-of-00044.safetensors",
820
+ "model.layers.72.mlp.gate_proj.weight": "model-00027-of-00044.safetensors",
821
+ "model.layers.72.mlp.up_proj.weight": "model-00027-of-00044.safetensors",
822
+ "model.layers.72.post_attention_layernorm.weight": "model-00027-of-00044.safetensors",
823
+ "model.layers.72.self_attn.k_proj.weight": "model-00027-of-00044.safetensors",
824
+ "model.layers.72.self_attn.o_proj.weight": "model-00027-of-00044.safetensors",
825
+ "model.layers.72.self_attn.q_proj.weight": "model-00027-of-00044.safetensors",
826
+ "model.layers.72.self_attn.v_proj.weight": "model-00027-of-00044.safetensors",
827
+ "model.layers.73.input_layernorm.weight": "model-00027-of-00044.safetensors",
828
+ "model.layers.73.mlp.down_proj.weight": "model-00027-of-00044.safetensors",
829
+ "model.layers.73.mlp.gate_proj.weight": "model-00027-of-00044.safetensors",
830
+ "model.layers.73.mlp.up_proj.weight": "model-00027-of-00044.safetensors",
831
+ "model.layers.73.post_attention_layernorm.weight": "model-00027-of-00044.safetensors",
832
+ "model.layers.73.self_attn.k_proj.weight": "model-00027-of-00044.safetensors",
833
+ "model.layers.73.self_attn.o_proj.weight": "model-00027-of-00044.safetensors",
834
+ "model.layers.73.self_attn.q_proj.weight": "model-00027-of-00044.safetensors",
835
+ "model.layers.73.self_attn.v_proj.weight": "model-00027-of-00044.safetensors",
836
+ "model.layers.74.input_layernorm.weight": "model-00028-of-00044.safetensors",
837
+ "model.layers.74.mlp.down_proj.weight": "model-00028-of-00044.safetensors",
838
+ "model.layers.74.mlp.gate_proj.weight": "model-00028-of-00044.safetensors",
839
+ "model.layers.74.mlp.up_proj.weight": "model-00028-of-00044.safetensors",
840
+ "model.layers.74.post_attention_layernorm.weight": "model-00028-of-00044.safetensors",
841
+ "model.layers.74.self_attn.k_proj.weight": "model-00027-of-00044.safetensors",
842
+ "model.layers.74.self_attn.o_proj.weight": "model-00027-of-00044.safetensors",
843
+ "model.layers.74.self_attn.q_proj.weight": "model-00027-of-00044.safetensors",
844
+ "model.layers.74.self_attn.v_proj.weight": "model-00027-of-00044.safetensors",
845
+ "model.layers.75.input_layernorm.weight": "model-00028-of-00044.safetensors",
846
+ "model.layers.75.mlp.down_proj.weight": "model-00028-of-00044.safetensors",
847
+ "model.layers.75.mlp.gate_proj.weight": "model-00028-of-00044.safetensors",
848
+ "model.layers.75.mlp.up_proj.weight": "model-00028-of-00044.safetensors",
849
+ "model.layers.75.post_attention_layernorm.weight": "model-00028-of-00044.safetensors",
850
+ "model.layers.75.self_attn.k_proj.weight": "model-00028-of-00044.safetensors",
851
+ "model.layers.75.self_attn.o_proj.weight": "model-00028-of-00044.safetensors",
852
+ "model.layers.75.self_attn.q_proj.weight": "model-00028-of-00044.safetensors",
853
+ "model.layers.75.self_attn.v_proj.weight": "model-00028-of-00044.safetensors",
854
+ "model.layers.76.input_layernorm.weight": "model-00028-of-00044.safetensors",
855
+ "model.layers.76.mlp.down_proj.weight": "model-00028-of-00044.safetensors",
856
+ "model.layers.76.mlp.gate_proj.weight": "model-00028-of-00044.safetensors",
857
+ "model.layers.76.mlp.up_proj.weight": "model-00028-of-00044.safetensors",
858
+ "model.layers.76.post_attention_layernorm.weight": "model-00028-of-00044.safetensors",
859
+ "model.layers.76.self_attn.k_proj.weight": "model-00028-of-00044.safetensors",
860
+ "model.layers.76.self_attn.o_proj.weight": "model-00028-of-00044.safetensors",
861
+ "model.layers.76.self_attn.q_proj.weight": "model-00028-of-00044.safetensors",
862
+ "model.layers.76.self_attn.v_proj.weight": "model-00028-of-00044.safetensors",
863
+ "model.layers.77.input_layernorm.weight": "model-00029-of-00044.safetensors",
864
+ "model.layers.77.mlp.down_proj.weight": "model-00029-of-00044.safetensors",
865
+ "model.layers.77.mlp.gate_proj.weight": "model-00029-of-00044.safetensors",
866
+ "model.layers.77.mlp.up_proj.weight": "model-00029-of-00044.safetensors",
867
+ "model.layers.77.post_attention_layernorm.weight": "model-00029-of-00044.safetensors",
868
+ "model.layers.77.self_attn.k_proj.weight": "model-00028-of-00044.safetensors",
869
+ "model.layers.77.self_attn.o_proj.weight": "model-00029-of-00044.safetensors",
870
+ "model.layers.77.self_attn.q_proj.weight": "model-00028-of-00044.safetensors",
871
+ "model.layers.77.self_attn.v_proj.weight": "model-00028-of-00044.safetensors",
872
+ "model.layers.78.input_layernorm.weight": "model-00029-of-00044.safetensors",
873
+ "model.layers.78.mlp.down_proj.weight": "model-00029-of-00044.safetensors",
874
+ "model.layers.78.mlp.gate_proj.weight": "model-00029-of-00044.safetensors",
875
+ "model.layers.78.mlp.up_proj.weight": "model-00029-of-00044.safetensors",
876
+ "model.layers.78.post_attention_layernorm.weight": "model-00029-of-00044.safetensors",
877
+ "model.layers.78.self_attn.k_proj.weight": "model-00029-of-00044.safetensors",
878
+ "model.layers.78.self_attn.o_proj.weight": "model-00029-of-00044.safetensors",
879
+ "model.layers.78.self_attn.q_proj.weight": "model-00029-of-00044.safetensors",
880
+ "model.layers.78.self_attn.v_proj.weight": "model-00029-of-00044.safetensors",
881
+ "model.layers.79.input_layernorm.weight": "model-00029-of-00044.safetensors",
882
+ "model.layers.79.mlp.down_proj.weight": "model-00029-of-00044.safetensors",
883
+ "model.layers.79.mlp.gate_proj.weight": "model-00029-of-00044.safetensors",
884
+ "model.layers.79.mlp.up_proj.weight": "model-00029-of-00044.safetensors",
885
+ "model.layers.79.post_attention_layernorm.weight": "model-00029-of-00044.safetensors",
886
+ "model.layers.79.self_attn.k_proj.weight": "model-00029-of-00044.safetensors",
887
+ "model.layers.79.self_attn.o_proj.weight": "model-00029-of-00044.safetensors",
888
+ "model.layers.79.self_attn.q_proj.weight": "model-00029-of-00044.safetensors",
889
+ "model.layers.79.self_attn.v_proj.weight": "model-00029-of-00044.safetensors",
890
+ "model.layers.8.input_layernorm.weight": "model-00004-of-00044.safetensors",
891
+ "model.layers.8.mlp.down_proj.weight": "model-00004-of-00044.safetensors",
892
+ "model.layers.8.mlp.gate_proj.weight": "model-00004-of-00044.safetensors",
893
+ "model.layers.8.mlp.up_proj.weight": "model-00004-of-00044.safetensors",
894
+ "model.layers.8.post_attention_layernorm.weight": "model-00004-of-00044.safetensors",
895
+ "model.layers.8.self_attn.k_proj.weight": "model-00004-of-00044.safetensors",
896
+ "model.layers.8.self_attn.o_proj.weight": "model-00004-of-00044.safetensors",
897
+ "model.layers.8.self_attn.q_proj.weight": "model-00004-of-00044.safetensors",
898
+ "model.layers.8.self_attn.v_proj.weight": "model-00004-of-00044.safetensors",
899
+ "model.layers.80.input_layernorm.weight": "model-00030-of-00044.safetensors",
900
+ "model.layers.80.mlp.down_proj.weight": "model-00030-of-00044.safetensors",
901
+ "model.layers.80.mlp.gate_proj.weight": "model-00030-of-00044.safetensors",
902
+ "model.layers.80.mlp.up_proj.weight": "model-00030-of-00044.safetensors",
903
+ "model.layers.80.post_attention_layernorm.weight": "model-00030-of-00044.safetensors",
904
+ "model.layers.80.self_attn.k_proj.weight": "model-00030-of-00044.safetensors",
905
+ "model.layers.80.self_attn.o_proj.weight": "model-00030-of-00044.safetensors",
906
+ "model.layers.80.self_attn.q_proj.weight": "model-00030-of-00044.safetensors",
907
+ "model.layers.80.self_attn.v_proj.weight": "model-00030-of-00044.safetensors",
908
+ "model.layers.81.input_layernorm.weight": "model-00030-of-00044.safetensors",
909
+ "model.layers.81.mlp.down_proj.weight": "model-00030-of-00044.safetensors",
910
+ "model.layers.81.mlp.gate_proj.weight": "model-00030-of-00044.safetensors",
911
+ "model.layers.81.mlp.up_proj.weight": "model-00030-of-00044.safetensors",
912
+ "model.layers.81.post_attention_layernorm.weight": "model-00030-of-00044.safetensors",
913
+ "model.layers.81.self_attn.k_proj.weight": "model-00030-of-00044.safetensors",
914
+ "model.layers.81.self_attn.o_proj.weight": "model-00030-of-00044.safetensors",
915
+ "model.layers.81.self_attn.q_proj.weight": "model-00030-of-00044.safetensors",
916
+ "model.layers.81.self_attn.v_proj.weight": "model-00030-of-00044.safetensors",
917
+ "model.layers.82.input_layernorm.weight": "model-00031-of-00044.safetensors",
918
+ "model.layers.82.mlp.down_proj.weight": "model-00031-of-00044.safetensors",
919
+ "model.layers.82.mlp.gate_proj.weight": "model-00030-of-00044.safetensors",
920
+ "model.layers.82.mlp.up_proj.weight": "model-00030-of-00044.safetensors",
921
+ "model.layers.82.post_attention_layernorm.weight": "model-00031-of-00044.safetensors",
922
+ "model.layers.82.self_attn.k_proj.weight": "model-00030-of-00044.safetensors",
923
+ "model.layers.82.self_attn.o_proj.weight": "model-00030-of-00044.safetensors",
924
+ "model.layers.82.self_attn.q_proj.weight": "model-00030-of-00044.safetensors",
925
+ "model.layers.82.self_attn.v_proj.weight": "model-00030-of-00044.safetensors",
926
+ "model.layers.83.input_layernorm.weight": "model-00031-of-00044.safetensors",
927
+ "model.layers.83.mlp.down_proj.weight": "model-00031-of-00044.safetensors",
928
+ "model.layers.83.mlp.gate_proj.weight": "model-00031-of-00044.safetensors",
929
+ "model.layers.83.mlp.up_proj.weight": "model-00031-of-00044.safetensors",
930
+ "model.layers.83.post_attention_layernorm.weight": "model-00031-of-00044.safetensors",
931
+ "model.layers.83.self_attn.k_proj.weight": "model-00031-of-00044.safetensors",
932
+ "model.layers.83.self_attn.o_proj.weight": "model-00031-of-00044.safetensors",
933
+ "model.layers.83.self_attn.q_proj.weight": "model-00031-of-00044.safetensors",
934
+ "model.layers.83.self_attn.v_proj.weight": "model-00031-of-00044.safetensors",
935
+ "model.layers.84.input_layernorm.weight": "model-00031-of-00044.safetensors",
936
+ "model.layers.84.mlp.down_proj.weight": "model-00031-of-00044.safetensors",
937
+ "model.layers.84.mlp.gate_proj.weight": "model-00031-of-00044.safetensors",
938
+ "model.layers.84.mlp.up_proj.weight": "model-00031-of-00044.safetensors",
939
+ "model.layers.84.post_attention_layernorm.weight": "model-00031-of-00044.safetensors",
940
+ "model.layers.84.self_attn.k_proj.weight": "model-00031-of-00044.safetensors",
941
+ "model.layers.84.self_attn.o_proj.weight": "model-00031-of-00044.safetensors",
942
+ "model.layers.84.self_attn.q_proj.weight": "model-00031-of-00044.safetensors",
943
+ "model.layers.84.self_attn.v_proj.weight": "model-00031-of-00044.safetensors",
944
+ "model.layers.85.input_layernorm.weight": "model-00032-of-00044.safetensors",
945
+ "model.layers.85.mlp.down_proj.weight": "model-00032-of-00044.safetensors",
946
+ "model.layers.85.mlp.gate_proj.weight": "model-00031-of-00044.safetensors",
947
+ "model.layers.85.mlp.up_proj.weight": "model-00032-of-00044.safetensors",
948
+ "model.layers.85.post_attention_layernorm.weight": "model-00032-of-00044.safetensors",
949
+ "model.layers.85.self_attn.k_proj.weight": "model-00031-of-00044.safetensors",
950
+ "model.layers.85.self_attn.o_proj.weight": "model-00031-of-00044.safetensors",
951
+ "model.layers.85.self_attn.q_proj.weight": "model-00031-of-00044.safetensors",
952
+ "model.layers.85.self_attn.v_proj.weight": "model-00031-of-00044.safetensors",
953
+ "model.layers.86.input_layernorm.weight": "model-00032-of-00044.safetensors",
954
+ "model.layers.86.mlp.down_proj.weight": "model-00032-of-00044.safetensors",
955
+ "model.layers.86.mlp.gate_proj.weight": "model-00032-of-00044.safetensors",
956
+ "model.layers.86.mlp.up_proj.weight": "model-00032-of-00044.safetensors",
957
+ "model.layers.86.post_attention_layernorm.weight": "model-00032-of-00044.safetensors",
958
+ "model.layers.86.self_attn.k_proj.weight": "model-00032-of-00044.safetensors",
959
+ "model.layers.86.self_attn.o_proj.weight": "model-00032-of-00044.safetensors",
960
+ "model.layers.86.self_attn.q_proj.weight": "model-00032-of-00044.safetensors",
961
+ "model.layers.86.self_attn.v_proj.weight": "model-00032-of-00044.safetensors",
962
+ "model.layers.87.input_layernorm.weight": "model-00032-of-00044.safetensors",
963
+ "model.layers.87.mlp.down_proj.weight": "model-00032-of-00044.safetensors",
964
+ "model.layers.87.mlp.gate_proj.weight": "model-00032-of-00044.safetensors",
965
+ "model.layers.87.mlp.up_proj.weight": "model-00032-of-00044.safetensors",
966
+ "model.layers.87.post_attention_layernorm.weight": "model-00032-of-00044.safetensors",
967
+ "model.layers.87.self_attn.k_proj.weight": "model-00032-of-00044.safetensors",
968
+ "model.layers.87.self_attn.o_proj.weight": "model-00032-of-00044.safetensors",
969
+ "model.layers.87.self_attn.q_proj.weight": "model-00032-of-00044.safetensors",
970
+ "model.layers.87.self_attn.v_proj.weight": "model-00032-of-00044.safetensors",
971
+ "model.layers.88.input_layernorm.weight": "model-00033-of-00044.safetensors",
972
+ "model.layers.88.mlp.down_proj.weight": "model-00033-of-00044.safetensors",
973
+ "model.layers.88.mlp.gate_proj.weight": "model-00033-of-00044.safetensors",
974
+ "model.layers.88.mlp.up_proj.weight": "model-00033-of-00044.safetensors",
975
+ "model.layers.88.post_attention_layernorm.weight": "model-00033-of-00044.safetensors",
976
+ "model.layers.88.self_attn.k_proj.weight": "model-00032-of-00044.safetensors",
977
+ "model.layers.88.self_attn.o_proj.weight": "model-00032-of-00044.safetensors",
978
+ "model.layers.88.self_attn.q_proj.weight": "model-00032-of-00044.safetensors",
979
+ "model.layers.88.self_attn.v_proj.weight": "model-00032-of-00044.safetensors",
980
+ "model.layers.89.input_layernorm.weight": "model-00033-of-00044.safetensors",
981
+ "model.layers.89.mlp.down_proj.weight": "model-00033-of-00044.safetensors",
982
+ "model.layers.89.mlp.gate_proj.weight": "model-00033-of-00044.safetensors",
983
+ "model.layers.89.mlp.up_proj.weight": "model-00033-of-00044.safetensors",
984
+ "model.layers.89.post_attention_layernorm.weight": "model-00033-of-00044.safetensors",
985
+ "model.layers.89.self_attn.k_proj.weight": "model-00033-of-00044.safetensors",
986
+ "model.layers.89.self_attn.o_proj.weight": "model-00033-of-00044.safetensors",
987
+ "model.layers.89.self_attn.q_proj.weight": "model-00033-of-00044.safetensors",
988
+ "model.layers.89.self_attn.v_proj.weight": "model-00033-of-00044.safetensors",
989
+ "model.layers.9.input_layernorm.weight": "model-00004-of-00044.safetensors",
990
+ "model.layers.9.mlp.down_proj.weight": "model-00004-of-00044.safetensors",
991
+ "model.layers.9.mlp.gate_proj.weight": "model-00004-of-00044.safetensors",
992
+ "model.layers.9.mlp.up_proj.weight": "model-00004-of-00044.safetensors",
993
+ "model.layers.9.post_attention_layernorm.weight": "model-00004-of-00044.safetensors",
994
+ "model.layers.9.self_attn.k_proj.weight": "model-00004-of-00044.safetensors",
995
+ "model.layers.9.self_attn.o_proj.weight": "model-00004-of-00044.safetensors",
996
+ "model.layers.9.self_attn.q_proj.weight": "model-00004-of-00044.safetensors",
997
+ "model.layers.9.self_attn.v_proj.weight": "model-00004-of-00044.safetensors",
998
+ "model.layers.90.input_layernorm.weight": "model-00033-of-00044.safetensors",
999
+ "model.layers.90.mlp.down_proj.weight": "model-00033-of-00044.safetensors",
1000
+ "model.layers.90.mlp.gate_proj.weight": "model-00033-of-00044.safetensors",
1001
+ "model.layers.90.mlp.up_proj.weight": "model-00033-of-00044.safetensors",
1002
+ "model.layers.90.post_attention_layernorm.weight": "model-00033-of-00044.safetensors",
1003
+ "model.layers.90.self_attn.k_proj.weight": "model-00033-of-00044.safetensors",
1004
+ "model.layers.90.self_attn.o_proj.weight": "model-00033-of-00044.safetensors",
1005
+ "model.layers.90.self_attn.q_proj.weight": "model-00033-of-00044.safetensors",
1006
+ "model.layers.90.self_attn.v_proj.weight": "model-00033-of-00044.safetensors",
1007
+ "model.layers.91.input_layernorm.weight": "model-00034-of-00044.safetensors",
1008
+ "model.layers.91.mlp.down_proj.weight": "model-00034-of-00044.safetensors",
1009
+ "model.layers.91.mlp.gate_proj.weight": "model-00034-of-00044.safetensors",
1010
+ "model.layers.91.mlp.up_proj.weight": "model-00034-of-00044.safetensors",
1011
+ "model.layers.91.post_attention_layernorm.weight": "model-00034-of-00044.safetensors",
1012
+ "model.layers.91.self_attn.k_proj.weight": "model-00033-of-00044.safetensors",
1013
+ "model.layers.91.self_attn.o_proj.weight": "model-00034-of-00044.safetensors",
1014
+ "model.layers.91.self_attn.q_proj.weight": "model-00033-of-00044.safetensors",
1015
+ "model.layers.91.self_attn.v_proj.weight": "model-00033-of-00044.safetensors",
1016
+ "model.layers.92.input_layernorm.weight": "model-00034-of-00044.safetensors",
1017
+ "model.layers.92.mlp.down_proj.weight": "model-00034-of-00044.safetensors",
1018
+ "model.layers.92.mlp.gate_proj.weight": "model-00034-of-00044.safetensors",
1019
+ "model.layers.92.mlp.up_proj.weight": "model-00034-of-00044.safetensors",
1020
+ "model.layers.92.post_attention_layernorm.weight": "model-00034-of-00044.safetensors",
1021
+ "model.layers.92.self_attn.k_proj.weight": "model-00034-of-00044.safetensors",
1022
+ "model.layers.92.self_attn.o_proj.weight": "model-00034-of-00044.safetensors",
1023
+ "model.layers.92.self_attn.q_proj.weight": "model-00034-of-00044.safetensors",
1024
+ "model.layers.92.self_attn.v_proj.weight": "model-00034-of-00044.safetensors",
1025
+ "model.layers.93.input_layernorm.weight": "model-00034-of-00044.safetensors",
1026
+ "model.layers.93.mlp.down_proj.weight": "model-00034-of-00044.safetensors",
1027
+ "model.layers.93.mlp.gate_proj.weight": "model-00034-of-00044.safetensors",
1028
+ "model.layers.93.mlp.up_proj.weight": "model-00034-of-00044.safetensors",
1029
+ "model.layers.93.post_attention_layernorm.weight": "model-00034-of-00044.safetensors",
1030
+ "model.layers.93.self_attn.k_proj.weight": "model-00034-of-00044.safetensors",
1031
+ "model.layers.93.self_attn.o_proj.weight": "model-00034-of-00044.safetensors",
1032
+ "model.layers.93.self_attn.q_proj.weight": "model-00034-of-00044.safetensors",
1033
+ "model.layers.93.self_attn.v_proj.weight": "model-00034-of-00044.safetensors",
1034
+ "model.layers.94.input_layernorm.weight": "model-00035-of-00044.safetensors",
1035
+ "model.layers.94.mlp.down_proj.weight": "model-00035-of-00044.safetensors",
1036
+ "model.layers.94.mlp.gate_proj.weight": "model-00035-of-00044.safetensors",
1037
+ "model.layers.94.mlp.up_proj.weight": "model-00035-of-00044.safetensors",
1038
+ "model.layers.94.post_attention_layernorm.weight": "model-00035-of-00044.safetensors",
1039
+ "model.layers.94.self_attn.k_proj.weight": "model-00035-of-00044.safetensors",
1040
+ "model.layers.94.self_attn.o_proj.weight": "model-00035-of-00044.safetensors",
1041
+ "model.layers.94.self_attn.q_proj.weight": "model-00035-of-00044.safetensors",
1042
+ "model.layers.94.self_attn.v_proj.weight": "model-00035-of-00044.safetensors",
1043
+ "model.layers.95.input_layernorm.weight": "model-00035-of-00044.safetensors",
1044
+ "model.layers.95.mlp.down_proj.weight": "model-00035-of-00044.safetensors",
1045
+ "model.layers.95.mlp.gate_proj.weight": "model-00035-of-00044.safetensors",
1046
+ "model.layers.95.mlp.up_proj.weight": "model-00035-of-00044.safetensors",
1047
+ "model.layers.95.post_attention_layernorm.weight": "model-00035-of-00044.safetensors",
1048
+ "model.layers.95.self_attn.k_proj.weight": "model-00035-of-00044.safetensors",
1049
+ "model.layers.95.self_attn.o_proj.weight": "model-00035-of-00044.safetensors",
1050
+ "model.layers.95.self_attn.q_proj.weight": "model-00035-of-00044.safetensors",
1051
+ "model.layers.95.self_attn.v_proj.weight": "model-00035-of-00044.safetensors",
1052
+ "model.layers.96.input_layernorm.weight": "model-00036-of-00044.safetensors",
1053
+ "model.layers.96.mlp.down_proj.weight": "model-00036-of-00044.safetensors",
1054
+ "model.layers.96.mlp.gate_proj.weight": "model-00035-of-00044.safetensors",
1055
+ "model.layers.96.mlp.up_proj.weight": "model-00035-of-00044.safetensors",
1056
+ "model.layers.96.post_attention_layernorm.weight": "model-00036-of-00044.safetensors",
1057
+ "model.layers.96.self_attn.k_proj.weight": "model-00035-of-00044.safetensors",
1058
+ "model.layers.96.self_attn.o_proj.weight": "model-00035-of-00044.safetensors",
1059
+ "model.layers.96.self_attn.q_proj.weight": "model-00035-of-00044.safetensors",
1060
+ "model.layers.96.self_attn.v_proj.weight": "model-00035-of-00044.safetensors",
1061
+ "model.layers.97.input_layernorm.weight": "model-00036-of-00044.safetensors",
1062
+ "model.layers.97.mlp.down_proj.weight": "model-00036-of-00044.safetensors",
1063
+ "model.layers.97.mlp.gate_proj.weight": "model-00036-of-00044.safetensors",
1064
+ "model.layers.97.mlp.up_proj.weight": "model-00036-of-00044.safetensors",
1065
+ "model.layers.97.post_attention_layernorm.weight": "model-00036-of-00044.safetensors",
1066
+ "model.layers.97.self_attn.k_proj.weight": "model-00036-of-00044.safetensors",
1067
+ "model.layers.97.self_attn.o_proj.weight": "model-00036-of-00044.safetensors",
1068
+ "model.layers.97.self_attn.q_proj.weight": "model-00036-of-00044.safetensors",
1069
+ "model.layers.97.self_attn.v_proj.weight": "model-00036-of-00044.safetensors",
1070
+ "model.layers.98.input_layernorm.weight": "model-00036-of-00044.safetensors",
1071
+ "model.layers.98.mlp.down_proj.weight": "model-00036-of-00044.safetensors",
1072
+ "model.layers.98.mlp.gate_proj.weight": "model-00036-of-00044.safetensors",
1073
+ "model.layers.98.mlp.up_proj.weight": "model-00036-of-00044.safetensors",
1074
+ "model.layers.98.post_attention_layernorm.weight": "model-00036-of-00044.safetensors",
1075
+ "model.layers.98.self_attn.k_proj.weight": "model-00036-of-00044.safetensors",
1076
+ "model.layers.98.self_attn.o_proj.weight": "model-00036-of-00044.safetensors",
1077
+ "model.layers.98.self_attn.q_proj.weight": "model-00036-of-00044.safetensors",
1078
+ "model.layers.98.self_attn.v_proj.weight": "model-00036-of-00044.safetensors",
1079
+ "model.layers.99.input_layernorm.weight": "model-00037-of-00044.safetensors",
1080
+ "model.layers.99.mlp.down_proj.weight": "model-00037-of-00044.safetensors",
1081
+ "model.layers.99.mlp.gate_proj.weight": "model-00036-of-00044.safetensors",
1082
+ "model.layers.99.mlp.up_proj.weight": "model-00037-of-00044.safetensors",
1083
+ "model.layers.99.post_attention_layernorm.weight": "model-00037-of-00044.safetensors",
1084
+ "model.layers.99.self_attn.k_proj.weight": "model-00036-of-00044.safetensors",
1085
+ "model.layers.99.self_attn.o_proj.weight": "model-00036-of-00044.safetensors",
1086
+ "model.layers.99.self_attn.q_proj.weight": "model-00036-of-00044.safetensors",
1087
+ "model.layers.99.self_attn.v_proj.weight": "model-00036-of-00044.safetensors",
1088
+ "model.norm.weight": "model-00044-of-00044.safetensors"
1089
+ }
1090
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|begin_of_text|>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<|eot_id|>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "<|finetune_right_pad_id|>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ }
23
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b
3
+ size 17209920