Bturtel commited on
Commit
b92728f
·
verified ·
1 Parent(s): df86c01

Remove merged model files (replaced with LoRA adapter)

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. config.json +0 -80
  2. model-00000-of-00014.safetensors +0 -3
  3. model-00001-of-00014.safetensors +0 -3
  4. model-00001-of-00073.safetensors +0 -3
  5. model-00002-of-00014.safetensors +0 -3
  6. model-00002-of-00073.safetensors +0 -3
  7. model-00003-of-00014.safetensors +0 -3
  8. model-00003-of-00073.safetensors +0 -3
  9. model-00004-of-00014.safetensors +0 -3
  10. model-00004-of-00073.safetensors +0 -3
  11. model-00005-of-00014.safetensors +0 -3
  12. model-00005-of-00073.safetensors +0 -3
  13. model-00006-of-00014.safetensors +0 -3
  14. model-00006-of-00073.safetensors +0 -3
  15. model-00007-of-00014.safetensors +0 -3
  16. model-00007-of-00073.safetensors +0 -3
  17. model-00008-of-00014.safetensors +0 -3
  18. model-00008-of-00073.safetensors +0 -3
  19. model-00009-of-00014.safetensors +0 -3
  20. model-00009-of-00073.safetensors +0 -3
  21. model-00010-of-00014.safetensors +0 -3
  22. model-00010-of-00073.safetensors +0 -3
  23. model-00011-of-00014.safetensors +0 -3
  24. model-00011-of-00073.safetensors +0 -3
  25. model-00012-of-00014.safetensors +0 -3
  26. model-00012-of-00073.safetensors +0 -3
  27. model-00013-of-00014.safetensors +0 -3
  28. model-00013-of-00073.safetensors +0 -3
  29. model-00014-of-00014.safetensors +0 -3
  30. model-00014-of-00073.safetensors +0 -3
  31. model-00015-of-00073.safetensors +0 -3
  32. model-00016-of-00073.safetensors +0 -3
  33. model-00017-of-00073.safetensors +0 -3
  34. model-00018-of-00073.safetensors +0 -3
  35. model-00019-of-00073.safetensors +0 -3
  36. model-00020-of-00073.safetensors +0 -3
  37. model-00021-of-00073.safetensors +0 -3
  38. model-00022-of-00073.safetensors +0 -3
  39. model-00023-of-00073.safetensors +0 -3
  40. model-00024-of-00073.safetensors +0 -3
  41. model-00025-of-00073.safetensors +0 -3
  42. model-00026-of-00073.safetensors +0 -3
  43. model-00027-of-00073.safetensors +0 -3
  44. model-00028-of-00073.safetensors +0 -3
  45. model-00029-of-00073.safetensors +0 -3
  46. model-00030-of-00073.safetensors +0 -3
  47. model-00031-of-00073.safetensors +0 -3
  48. model-00032-of-00073.safetensors +0 -3
  49. model-00033-of-00073.safetensors +0 -3
  50. model-00034-of-00073.safetensors +0 -3
config.json DELETED
@@ -1,80 +0,0 @@
1
- {
2
- "architectures": [
3
- "GptOssForCausalLM"
4
- ],
5
- "attention_bias": true,
6
- "attention_dropout": 0.0,
7
- "eos_token_id": 200002,
8
- "experts_per_token": 4,
9
- "head_dim": 64,
10
- "hidden_act": "silu",
11
- "hidden_size": 2880,
12
- "initial_context_length": 4096,
13
- "initializer_range": 0.02,
14
- "intermediate_size": 2880,
15
- "layer_types": [
16
- "sliding_attention",
17
- "full_attention",
18
- "sliding_attention",
19
- "full_attention",
20
- "sliding_attention",
21
- "full_attention",
22
- "sliding_attention",
23
- "full_attention",
24
- "sliding_attention",
25
- "full_attention",
26
- "sliding_attention",
27
- "full_attention",
28
- "sliding_attention",
29
- "full_attention",
30
- "sliding_attention",
31
- "full_attention",
32
- "sliding_attention",
33
- "full_attention",
34
- "sliding_attention",
35
- "full_attention",
36
- "sliding_attention",
37
- "full_attention",
38
- "sliding_attention",
39
- "full_attention",
40
- "sliding_attention",
41
- "full_attention",
42
- "sliding_attention",
43
- "full_attention",
44
- "sliding_attention",
45
- "full_attention",
46
- "sliding_attention",
47
- "full_attention",
48
- "sliding_attention",
49
- "full_attention",
50
- "sliding_attention",
51
- "full_attention"
52
- ],
53
- "max_position_embeddings": 131072,
54
- "model_type": "gpt_oss",
55
- "num_attention_heads": 64,
56
- "num_experts_per_tok": 4,
57
- "num_hidden_layers": 36,
58
- "num_key_value_heads": 8,
59
- "num_local_experts": 128,
60
- "output_router_logits": false,
61
- "pad_token_id": 199999,
62
- "rms_norm_eps": 1e-05,
63
- "rope_scaling": {
64
- "beta_fast": 32.0,
65
- "beta_slow": 1.0,
66
- "factor": 32.0,
67
- "original_max_position_embeddings": 4096,
68
- "rope_type": "yarn",
69
- "truncate": false
70
- },
71
- "rope_theta": 150000,
72
- "router_aux_loss_coef": 0.9,
73
- "sliding_window": 128,
74
- "swiglu_limit": 7.0,
75
- "tie_word_embeddings": false,
76
- "transformers_version": "4.55.0.dev0",
77
- "use_cache": true,
78
- "vocab_size": 201088,
79
- "torch_dtype": "bfloat16"
80
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
model-00000-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:695218884684c611fe08a74751ee443f971e9bd9bc062edba822da3fe45969b7
3
- size 4625017896
 
 
 
 
model-00001-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:a881aa5f561b26a22b14a8262aa61849ace349ffd73d74769e030ac90a1fcf8a
3
- size 4115586736
 
 
 
 
model-00001-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:3914b9e47635ff09120b12a8d1b9066ca9b3680100ca20f44647f9a4c18f0c74
3
- size 1212106032
 
 
 
 
model-00002-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:022478dd04398c5bdb545a5be0a6437ecc2eb53d1dbd29edafcfff4b3ddf0a41
3
- size 4625017888
 
 
 
 
model-00002-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:330a7853eafe265073156dbc3478fdf4c91be23e7927ac82fb66f041ea2caecc
3
- size 4248207608
 
 
 
 
model-00003-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:47aee9e7b9d5bedb215042c01ccededd9bd9c30b0dddea862dc2506b9d6c74de
3
- size 4115586752
 
 
 
 
model-00003-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:ccad652f94cd74f6d5571ad1cf9a4a89132e0acdb3a372939d11a919c30203e6
3
- size 2177954704
 
 
 
 
model-00004-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:f6c2752acda607b1d5ca52df9e75c1b9b2761e6875ff10c9bd6ddac473c0262e
3
- size 4625017896
 
 
 
 
model-00004-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:785b41a8fd53adbd0ca6448081a0e178eb1755fea67f5712e79aaa8aac326b7e
3
- size 4248207608
 
 
 
 
model-00005-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:0c8dd401544c31cb93b8459eee7da20ea2a07626a59455d7d92b85257df9b46c
3
- size 4115586696
 
 
 
 
model-00005-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c9c59e5882005748564a57d002cee9aa9bc0d7f47970cfeb2eedb859b5ad91f7
3
- size 2177954704
 
 
 
 
model-00006-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:28d839f2e027985a8b14e45f2323798862eddb7770ee9800ea6b7c803abee489
3
- size 4625017856
 
 
 
 
model-00006-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:d4afeef53609f547bbe4dfd1ba068671c302f056c41c5258c2f08e5718024fcd
3
- size 4248207608
 
 
 
 
model-00007-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c8958c5f183c04f6ea959cfd90562b5128124154b2bbf979b8a22b9405b30ed8
3
- size 4060267176
 
 
 
 
model-00007-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:d02dba1d5cf2ca878f46a11d51ee6cacb04cb911e0e58f9379469bdd84039302
3
- size 2177954704
 
 
 
 
model-00008-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:bf1f2a88868ffc37d520dcf77d26f0e823710b5e682d473ff10f6974fa3b7517
3
- size 4625017896
 
 
 
 
model-00008-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c6cb2f3ffff5bb15388d6b87cabffa8a08c2f6748b4186509c8859264fe6bae8
3
- size 4248207608
 
 
 
 
model-00009-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:f72d34a4004241b45c332b61f8ffa124e9a913bc1ab442b66e717d3e94e741ce
3
- size 4170906304
 
 
 
 
model-00009-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:12d8ea1c16fb77b608258658638aed138252fd7c8f61e2adb9d2d430d0daf925
3
- size 2177954704
 
 
 
 
model-00010-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:f48c867c2cb0a44bfc2f8768cb98e4aec9a350946fceacfebdcad5d32ad4a471
3
- size 4625017896
 
 
 
 
model-00010-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:d780df3aed5ab8d3e0ac2679f0e9e9c4ddec6c28edf521eebaddfb54a4c01c6e
3
- size 4248207608
 
 
 
 
model-00011-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:a06851b2cfd35f48722f823bc1ab8f7bcb4a878a5b8e975f4d3544f230454eeb
3
- size 4115586752
 
 
 
 
model-00011-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:d8e26a9cde5ade1d139621c8c4965bf5609a05a86690e352c0acd991f15bf39d
3
- size 2177954704
 
 
 
 
model-00012-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:3af33667c307e20ae2a7648ea52653de46dd0171601ec5c696e47a2f5d5bf1e4
3
- size 4064660808
 
 
 
 
model-00012-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:839223b330ca91a30f2af17520db90eae73618d2511ca684e9c5627f14dafb5c
3
- size 4248207608
 
 
 
 
model-00013-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c52f4cbe342e2cb0f1497ce059000973218b35c1e56816dd8683fe485bcdac82
3
- size 1902116864
 
 
 
 
model-00013-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c06b561df31023d8ab6ee29b6ae722cecbbeffe1690f28a9373f5dc9955c1bbe
3
- size 2177954704
 
 
 
 
model-00014-of-00014.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:54b1be1609696c307cc5ca117b1fa54feaddebffa04e9c2db117652a01964230
3
- size 4115586736
 
 
 
 
model-00014-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:071b6de5fe46555142d2ae18536fc5eadd1237c8ba3576d3fcb4ee2a488aa081
3
- size 4248207608
 
 
 
 
model-00015-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:f3f5c51935e3418191641055ebdacb97b7d0b368ac06b4852e40707ea59c2637
3
- size 2177954704
 
 
 
 
model-00016-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:e0416580a4650d6f6308a34a26e3bbae381abfcec97e0e0da30c79255be586f3
3
- size 4248207608
 
 
 
 
model-00017-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:ddcd373d64204ce8acea5d464df161a6b97c5762ef418db7a6a33b6d190688e0
3
- size 2177954704
 
 
 
 
model-00018-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:87c76cc955ef820987d4a1e038d9c0829a5feec2b09168be23869176f6d2c599
3
- size 4248207608
 
 
 
 
model-00019-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:09f12ba29a3e7827f700d5be3cf1250913babc9c8b4579c655cda949d7ec2302
3
- size 2177954704
 
 
 
 
model-00020-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:fc462f6dd2e3ca5b53f048469872bef103ca7421cb1cd99a36540b75dfa10529
3
- size 4248207608
 
 
 
 
model-00021-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:f9e7f7ed117a3cfa773647e22e50b18af95d6afdfd3b27ed342bbb18a95b85ef
3
- size 2177954656
 
 
 
 
model-00022-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:96dbeda00ec25ed81926014a97fc3fe4fc9b29251b211a5ca02adbd7669e2f2d
3
- size 4248207608
 
 
 
 
model-00023-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:69700f9a3a690350be808f3aae3993b294d3e504e3f09e445960ced149bb6cf7
3
- size 2177954720
 
 
 
 
model-00024-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:5c7c5cc68dd9bd22739bfc1241905466c9f94223cdd37c75df56e2b52fbb1433
3
- size 4248207608
 
 
 
 
model-00025-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:3e169503316f450cba5ddb647672f49bf664b08ea630532647fa2074b2a4fd8b
3
- size 2177954720
 
 
 
 
model-00026-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:fab908aeec305e091da1378a7fdbd4a36cb133346639e3ae2de646885e775a92
3
- size 4248207608
 
 
 
 
model-00027-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:ddf5ed49b9d8175d5df4378d07046ae55e2918a80828c33beed47c6b24537ab6
3
- size 2177954720
 
 
 
 
model-00028-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:b38ae58be6b3cd695a31ed996a51e4b025fc18b0b3363d26a3035e4c9107f292
3
- size 4248207608
 
 
 
 
model-00029-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:182d1533641ca2171285f062211ad318b69632ea9e5c8e74d4aa1bbe6774a345
3
- size 2177954720
 
 
 
 
model-00030-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:801e9bba14f900af03acac078d9e2f25f56ba17a3df0cf58f497bd9dc717a3a5
3
- size 4248207608
 
 
 
 
model-00031-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:77253965d6dab34b78ac8409c1443d402caa9a0913bb0f461eb6f242a44c1de6
3
- size 2177954720
 
 
 
 
model-00032-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:bd4b467d698348f4167d8ce1f146571ce3a115bb34e72beb2513adde7e3c7013
3
- size 4248207608
 
 
 
 
model-00033-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:57a360500976d8d3f8397f954227dcd1ca00a892e1037a503cf2c71e888bd868
3
- size 2177954720
 
 
 
 
model-00034-of-00073.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:2d4500db25d1e0cd4c8af3b64d35c282c791e7a00fd20a806ca9eb08d0d5af4b
3
- size 4248207608