nielsbantilan commited on
Commit
30d3a01
1 Parent(s): fc7a881

Upload folder using huggingface_hub

Browse files
Files changed (45) hide show
  1. .gitattributes +9 -0
  2. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/config.json +26 -0
  3. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/debug-cli.root.log +0 -0
  4. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-15vedvr7/sandbox/local_flytekit/inputs.pb +3 -0
  5. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-2z_s3p01/sandbox/local_flytekit/06f72de872d28705906fd9f7a6ef1df7/00000 +3 -0
  6. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-3e_8c00_/sandbox/local_flytekit/9aa7c258f1e4fb940402ee71afacbbbb/00000 +3 -0
  7. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-jg2qqgrx/sandbox/local_flytekit/6b17fae8944179d014a2e72f9068bb17/00000 +3 -0
  8. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-l7do51st/sandbox/local_flytekit/1ece2271d60626a7c68509b2e35fa266/00000 +3 -0
  9. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-nwbtoq_z/sandbox/local_flytekit/899d2b63e67c7ebbf715a7861865dca8/00000 +3 -0
  10. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-pbs7il4k/sandbox/local_flytekit/755384ae37b2bef7bc07d212bee35c22/00000 +3 -0
  11. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-uruixz18/sandbox/local_flytekit/1525550a99b021eb4391c0302d22d55b/00000 +3 -0
  12. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-z4pk0owb/sandbox/local_flytekit/ff531e1dc85976e00f0b75473056a120/00000 +3 -0
  13. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte_training_config.json +1 -0
  14. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/generation_config.json +6 -0
  15. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/added_tokens.json +0 -0
  16. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/merges.txt +0 -0
  17. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/vocab.json +0 -0
  18. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/0204ed10c186a4c7c68f55dff8f26087a45898d6 +5 -0
  19. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/0204ed10c186a4c7c68f55dff8f26087a45898d6.lock +0 -0
  20. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/08015a39463145a657e88a093aad79df9b2a76bf +0 -0
  21. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/08015a39463145a657e88a093aad79df9b2a76bf.lock +0 -0
  22. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/344b8478e86f0cb7933a50ef1fff9127ed27266c +6 -0
  23. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/344b8478e86f0cb7933a50ef1fff9127ed27266c.lock +0 -0
  24. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/7cb72eb7357994a07d5c54ffdeef99dec84e21ae +25 -0
  25. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/7cb72eb7357994a07d5c54ffdeef99dec84e21ae.lock +0 -0
  26. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b +3 -0
  27. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b.lock +0 -0
  28. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/af28df4f9190240e1c3fb77a27b6f3d9bdd41c2d +9 -0
  29. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/refs/main +1 -0
  30. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/config.json +25 -0
  31. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/generation_config.json +6 -0
  32. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/pytorch_model.bin +3 -0
  33. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/special_tokens_map.json +5 -0
  34. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/tokenizer.json +0 -0
  35. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/tokenizer_config.json +9 -0
  36. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/pytorch_model.bin +3 -0
  37. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/special_tokens_map.json +6 -0
  38. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/tokenizer.json +0 -0
  39. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/tokenizer_config.json +11 -0
  40. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/trainer_state.json +25 -0
  41. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/training_args.bin +3 -0
  42. flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/version.txt +1 -0
  43. flyte_training_config.json +1 -1
  44. pytorch_model.bin +1 -1
  45. training_args.bin +1 -1
.gitattributes CHANGED
@@ -50,3 +50,12 @@ flytexhkrcgxw/local_flytekit/a0bdcc43a58af4ab83094e2ad7e965a6/flyte-nyxd1kb4/san
50
  flytexhkrcgxw/local_flytekit/a0bdcc43a58af4ab83094e2ad7e965a6/flyte-tkhkqwit/sandbox/local_flytekit/bda70de7ed2fa38e66f14a88bdc1b202/00000 filter=lfs diff=lfs merge=lfs -text
51
  flytexhkrcgxw/local_flytekit/a0bdcc43a58af4ab83094e2ad7e965a6/flyte-xu53uvea/sandbox/local_flytekit/ec9d782ca167a7ca4f577c815dabbbb5/00000 filter=lfs diff=lfs merge=lfs -text
52
  flytexhkrcgxw/local_flytekit/a0bdcc43a58af4ab83094e2ad7e965a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
50
  flytexhkrcgxw/local_flytekit/a0bdcc43a58af4ab83094e2ad7e965a6/flyte-tkhkqwit/sandbox/local_flytekit/bda70de7ed2fa38e66f14a88bdc1b202/00000 filter=lfs diff=lfs merge=lfs -text
51
  flytexhkrcgxw/local_flytekit/a0bdcc43a58af4ab83094e2ad7e965a6/flyte-xu53uvea/sandbox/local_flytekit/ec9d782ca167a7ca4f577c815dabbbb5/00000 filter=lfs diff=lfs merge=lfs -text
52
  flytexhkrcgxw/local_flytekit/a0bdcc43a58af4ab83094e2ad7e965a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b filter=lfs diff=lfs merge=lfs -text
53
+ flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-2z_s3p01/sandbox/local_flytekit/06f72de872d28705906fd9f7a6ef1df7/00000 filter=lfs diff=lfs merge=lfs -text
54
+ flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-3e_8c00_/sandbox/local_flytekit/9aa7c258f1e4fb940402ee71afacbbbb/00000 filter=lfs diff=lfs merge=lfs -text
55
+ flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-jg2qqgrx/sandbox/local_flytekit/6b17fae8944179d014a2e72f9068bb17/00000 filter=lfs diff=lfs merge=lfs -text
56
+ flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-l7do51st/sandbox/local_flytekit/1ece2271d60626a7c68509b2e35fa266/00000 filter=lfs diff=lfs merge=lfs -text
57
+ flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-nwbtoq_z/sandbox/local_flytekit/899d2b63e67c7ebbf715a7861865dca8/00000 filter=lfs diff=lfs merge=lfs -text
58
+ flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-pbs7il4k/sandbox/local_flytekit/755384ae37b2bef7bc07d212bee35c22/00000 filter=lfs diff=lfs merge=lfs -text
59
+ flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-uruixz18/sandbox/local_flytekit/1525550a99b021eb4391c0302d22d55b/00000 filter=lfs diff=lfs merge=lfs -text
60
+ flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-z4pk0owb/sandbox/local_flytekit/ff531e1dc85976e00f0b75473056a120/00000 filter=lfs diff=lfs merge=lfs -text
61
+ flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b filter=lfs diff=lfs merge=lfs -text
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "togethercomputer/RedPajama-INCITE-Base-3B-v1",
3
+ "architectures": [
4
+ "GPTNeoXForCausalLM"
5
+ ],
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": 0.1,
8
+ "eos_token_id": 0,
9
+ "hidden_act": "gelu",
10
+ "hidden_size": 2560,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 10240,
13
+ "layer_norm_eps": 1e-05,
14
+ "max_position_embeddings": 2048,
15
+ "model_type": "gpt_neox",
16
+ "num_attention_heads": 32,
17
+ "num_hidden_layers": 32,
18
+ "rotary_emb_base": 10000,
19
+ "rotary_pct": 1.0,
20
+ "tie_word_embeddings": false,
21
+ "torch_dtype": "float16",
22
+ "transformers_version": "4.29.2",
23
+ "use_cache": true,
24
+ "use_parallel_residual": false,
25
+ "vocab_size": 50432
26
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/debug-cli.root.log ADDED
File without changes
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-15vedvr7/sandbox/local_flytekit/inputs.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ceae04abcb6c94590b7d222c548a5386862b57fda47edf878503690cfd2c644
3
+ size 1839
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-2z_s3p01/sandbox/local_flytekit/06f72de872d28705906fd9f7a6ef1df7/00000 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
3
+ size 133886409
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-3e_8c00_/sandbox/local_flytekit/9aa7c258f1e4fb940402ee71afacbbbb/00000 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
3
+ size 133886409
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-jg2qqgrx/sandbox/local_flytekit/6b17fae8944179d014a2e72f9068bb17/00000 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
3
+ size 133886409
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-l7do51st/sandbox/local_flytekit/1ece2271d60626a7c68509b2e35fa266/00000 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
3
+ size 133886409
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-nwbtoq_z/sandbox/local_flytekit/899d2b63e67c7ebbf715a7861865dca8/00000 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
3
+ size 133886409
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-pbs7il4k/sandbox/local_flytekit/755384ae37b2bef7bc07d212bee35c22/00000 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
3
+ size 133886409
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-uruixz18/sandbox/local_flytekit/1525550a99b021eb4391c0302d22d55b/00000 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
3
+ size 133886409
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-z4pk0owb/sandbox/local_flytekit/ff531e1dc85976e00f0b75473056a120/00000 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
3
+ size 133886409
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte_training_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"base_model": "togethercomputer/RedPajama-INCITE-Base-3B-v1", "data_path": "wikipedia", "data_name": "20220301.simple", "num_epochs": 5, "max_steps": -1, "learning_rate": 2e-05, "weight_decay": 0.02, "warmup_ratio": 0.03, "lr_scheduler_type": "cosine", "batch_size": 16, "micro_batch_size": 1, "val_set_size": 0, "group_by_length": false, "instruction_key": "instruction", "input_key": "input", "output_key": "output", "device_map": "auto", "cache_dir": null, "optim": "adamw_torch", "model_max_length": 512, "debug_mode": false, "debug_train_data_size": 1024, "wandb_project": ""}
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "eos_token_id": 0,
5
+ "transformers_version": "4.29.2"
6
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/added_tokens.json ADDED
File without changes
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/merges.txt ADDED
File without changes
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/vocab.json ADDED
File without changes
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/0204ed10c186a4c7c68f55dff8f26087a45898d6 ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/0204ed10c186a4c7c68f55dff8f26087a45898d6.lock ADDED
File without changes
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/08015a39463145a657e88a093aad79df9b2a76bf ADDED
The diff for this file is too large to render. See raw diff
 
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/08015a39463145a657e88a093aad79df9b2a76bf.lock ADDED
File without changes
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/344b8478e86f0cb7933a50ef1fff9127ed27266c ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "eos_token_id": 0,
5
+ "transformers_version": "4.28.1"
6
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/344b8478e86f0cb7933a50ef1fff9127ed27266c.lock ADDED
File without changes
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/7cb72eb7357994a07d5c54ffdeef99dec84e21ae ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "rp_3b_800b",
3
+ "architectures": [
4
+ "GPTNeoXForCausalLM"
5
+ ],
6
+ "bos_token_id": 0,
7
+ "eos_token_id": 0,
8
+ "hidden_act": "gelu",
9
+ "hidden_size": 2560,
10
+ "initializer_range": 0.02,
11
+ "intermediate_size": 10240,
12
+ "layer_norm_eps": 1e-05,
13
+ "max_position_embeddings": 2048,
14
+ "model_type": "gpt_neox",
15
+ "num_attention_heads": 32,
16
+ "num_hidden_layers": 32,
17
+ "rotary_emb_base": 10000,
18
+ "rotary_pct": 1.0,
19
+ "tie_word_embeddings": false,
20
+ "torch_dtype": "float16",
21
+ "transformers_version": "4.28.1",
22
+ "use_cache": true,
23
+ "use_parallel_residual": false,
24
+ "vocab_size": 50432
25
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/7cb72eb7357994a07d5c54ffdeef99dec84e21ae.lock ADDED
File without changes
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b
3
+ size 5686113497
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b.lock ADDED
File without changes
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/af28df4f9190240e1c3fb77a27b6f3d9bdd41c2d ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "bos_token": "<|endoftext|>",
4
+ "clean_up_tokenization_spaces": true,
5
+ "eos_token": "<|endoftext|>",
6
+ "model_max_length": 2048,
7
+ "tokenizer_class": "GPTNeoXTokenizer",
8
+ "unk_token": "<|endoftext|>"
9
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/refs/main ADDED
@@ -0,0 +1 @@
 
 
1
+ 094fbdd0c911feb485ce55de1952ab2e75277e1e
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "rp_3b_800b",
3
+ "architectures": [
4
+ "GPTNeoXForCausalLM"
5
+ ],
6
+ "bos_token_id": 0,
7
+ "eos_token_id": 0,
8
+ "hidden_act": "gelu",
9
+ "hidden_size": 2560,
10
+ "initializer_range": 0.02,
11
+ "intermediate_size": 10240,
12
+ "layer_norm_eps": 1e-05,
13
+ "max_position_embeddings": 2048,
14
+ "model_type": "gpt_neox",
15
+ "num_attention_heads": 32,
16
+ "num_hidden_layers": 32,
17
+ "rotary_emb_base": 10000,
18
+ "rotary_pct": 1.0,
19
+ "tie_word_embeddings": false,
20
+ "torch_dtype": "float16",
21
+ "transformers_version": "4.28.1",
22
+ "use_cache": true,
23
+ "use_parallel_residual": false,
24
+ "vocab_size": 50432
25
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "eos_token_id": 0,
5
+ "transformers_version": "4.28.1"
6
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b
3
+ size 5686113497
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/tokenizer_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "bos_token": "<|endoftext|>",
4
+ "clean_up_tokenization_spaces": true,
5
+ "eos_token": "<|endoftext|>",
6
+ "model_max_length": 2048,
7
+ "tokenizer_class": "GPTNeoXTokenizer",
8
+ "unk_token": "<|endoftext|>"
9
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c3892be9a5a69f1bdc1a76ac3c79cf722f4608b39d7e72ba1ea4412996b3e0e
3
+ size 5686106713
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "pad_token": "[PAD]",
5
+ "unk_token": "<|endoftext|>"
6
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/tokenizer_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "bos_token": "<|endoftext|>",
4
+ "clean_up_tokenization_spaces": true,
5
+ "eos_token": "<|endoftext|>",
6
+ "model_max_length": 512,
7
+ "pad_token": "[PAD]",
8
+ "padding_side": "right",
9
+ "tokenizer_class": "GPTNeoXTokenizer",
10
+ "unk_token": "<|endoftext|>"
11
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/trainer_state.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.3333333333333335,
5
+ "global_step": 5,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 3.33,
12
+ "step": 5,
13
+ "total_flos": 2077564600320.0,
14
+ "train_loss": 2.06929931640625,
15
+ "train_runtime": 218.7837,
16
+ "train_samples_per_second": 4.319,
17
+ "train_steps_per_second": 0.023
18
+ }
19
+ ],
20
+ "max_steps": 5,
21
+ "num_train_epochs": 5,
22
+ "total_flos": 2077564600320.0,
23
+ "trial_name": null,
24
+ "trial_params": null
25
+ }
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f954b5220fecd13e835e992ca0ed4b4cbebb8a29156115f726c8b5905282a3a
3
+ size 5563
flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/version.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ 1
flyte_training_config.json CHANGED
@@ -1 +1 @@
1
- {"base_model": "togethercomputer/RedPajama-INCITE-Base-3B-v1", "data_path": "wikipedia", "data_name": "20220301.simple", "num_epochs": 3, "max_steps": -1, "learning_rate": 2e-05, "weight_decay": 0.02, "warmup_ratio": 0.03, "lr_scheduler_type": "cosine", "batch_size": 16, "micro_batch_size": 1, "val_set_size": 0, "group_by_length": false, "instruction_key": "instruction", "input_key": "input", "output_key": "output", "device_map": "auto", "cache_dir": null, "optim": "adamw_torch", "model_max_length": 512, "debug_mode": false, "debug_train_data_size": 1024, "wandb_project": ""}
 
1
+ {"base_model": "togethercomputer/RedPajama-INCITE-Base-3B-v1", "data_path": "wikipedia", "data_name": "20220301.simple", "num_epochs": 5, "max_steps": -1, "learning_rate": 2e-05, "weight_decay": 0.02, "warmup_ratio": 0.03, "lr_scheduler_type": "cosine", "batch_size": 16, "micro_batch_size": 1, "val_set_size": 0, "group_by_length": false, "instruction_key": "instruction", "input_key": "input", "output_key": "output", "device_map": "auto", "cache_dir": null, "optim": "adamw_torch", "model_max_length": 512, "debug_mode": false, "debug_train_data_size": 1024, "wandb_project": ""}
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5cf3264e3ed851419a0cd378b83a70dda4d7298add89b7b5d844e0cd890beb8f
3
  size 3169529757
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c7fed73f3ca052abf4965428545bffca4951d6cdd4d7bc1316e4eedb7d08637
3
  size 3169529757
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ee731ce9f4fb31d73fecfa97d3439645a74d7340661828ddc1b5632d1d41799e
3
  size 3899
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:08b08062463650c20488fac34e4e4253f42e5221a2d85fb32be0717ceb253039
3
  size 3899