nielsbantilan
commited on
Commit
•
b504890
1
Parent(s):
9fee59b
Upload folder using huggingface_hub
Browse files- .gitattributes +9 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/config.json +26 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/debug-cli.root.log +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-d3lu49g2/sandbox/local_flytekit/7202bf829fca4ce2b2d6f5eccbdca304/00000 +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-g0d6z68v/sandbox/local_flytekit/303f0bdbd1297f390becb36ed2b37a0a/00000 +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-g2fb47m8/sandbox/local_flytekit/e46090cfa982a6f7f7e471e7cc5452c7/00000 +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-kpzvlk9q/sandbox/local_flytekit/inputs.pb +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-lsafy50k/sandbox/local_flytekit/d656dfcde256ff740a41645b0a35573b/00000 +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-mkrrkkxt/sandbox/local_flytekit/d8175d827abaf1ea3da2c455259e80c6/00000 +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-rpzbckua/sandbox/local_flytekit/6cf5cd5b55d6331ecbd20837130884b8/00000 +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-so3w79m7/sandbox/local_flytekit/32f5ab2fae65d042c0a6d22025dea3ac/00000 +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-udcr2wl6/sandbox/local_flytekit/1e3263ea13800320141857b26ef32688/00000 +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte_training_config.json +1 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/generation_config.json +6 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/added_tokens.json +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/merges.txt +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/vocab.json +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/0204ed10c186a4c7c68f55dff8f26087a45898d6 +5 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/0204ed10c186a4c7c68f55dff8f26087a45898d6.lock +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/08015a39463145a657e88a093aad79df9b2a76bf +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/08015a39463145a657e88a093aad79df9b2a76bf.lock +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/344b8478e86f0cb7933a50ef1fff9127ed27266c +6 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/344b8478e86f0cb7933a50ef1fff9127ed27266c.lock +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/7cb72eb7357994a07d5c54ffdeef99dec84e21ae +25 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/7cb72eb7357994a07d5c54ffdeef99dec84e21ae.lock +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b.lock +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/af28df4f9190240e1c3fb77a27b6f3d9bdd41c2d +9 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/refs/main +1 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/config.json +25 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/generation_config.json +6 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/pytorch_model.bin +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/special_tokens_map.json +5 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/tokenizer.json +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/tokenizer_config.json +9 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/pytorch_model.bin +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/special_tokens_map.json +6 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/tokenizer.json +0 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/tokenizer_config.json +11 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/trainer_state.json +31 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/training_args.bin +3 -0
- flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/version.txt +1 -0
- flyte_training_config.json +1 -1
- pytorch_model.bin +1 -1
- training_args.bin +1 -1
.gitattributes
CHANGED
@@ -59,3 +59,12 @@ flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-pbs7il4k/san
|
|
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
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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
|
62 |
+
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-d3lu49g2/sandbox/local_flytekit/7202bf829fca4ce2b2d6f5eccbdca304/00000 filter=lfs diff=lfs merge=lfs -text
|
63 |
+
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-g0d6z68v/sandbox/local_flytekit/303f0bdbd1297f390becb36ed2b37a0a/00000 filter=lfs diff=lfs merge=lfs -text
|
64 |
+
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-g2fb47m8/sandbox/local_flytekit/e46090cfa982a6f7f7e471e7cc5452c7/00000 filter=lfs diff=lfs merge=lfs -text
|
65 |
+
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-lsafy50k/sandbox/local_flytekit/d656dfcde256ff740a41645b0a35573b/00000 filter=lfs diff=lfs merge=lfs -text
|
66 |
+
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-mkrrkkxt/sandbox/local_flytekit/d8175d827abaf1ea3da2c455259e80c6/00000 filter=lfs diff=lfs merge=lfs -text
|
67 |
+
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-rpzbckua/sandbox/local_flytekit/6cf5cd5b55d6331ecbd20837130884b8/00000 filter=lfs diff=lfs merge=lfs -text
|
68 |
+
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-so3w79m7/sandbox/local_flytekit/32f5ab2fae65d042c0a6d22025dea3ac/00000 filter=lfs diff=lfs merge=lfs -text
|
69 |
+
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-udcr2wl6/sandbox/local_flytekit/1e3263ea13800320141857b26ef32688/00000 filter=lfs diff=lfs merge=lfs -text
|
70 |
+
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b filter=lfs diff=lfs merge=lfs -text
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/debug-cli.root.log
ADDED
File without changes
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-d3lu49g2/sandbox/local_flytekit/7202bf829fca4ce2b2d6f5eccbdca304/00000
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
|
3 |
+
size 133886409
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-g0d6z68v/sandbox/local_flytekit/303f0bdbd1297f390becb36ed2b37a0a/00000
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
|
3 |
+
size 133886409
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-g2fb47m8/sandbox/local_flytekit/e46090cfa982a6f7f7e471e7cc5452c7/00000
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
|
3 |
+
size 133886409
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-kpzvlk9q/sandbox/local_flytekit/inputs.pb
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:db8a93ad392ab06c25eac142df7874c95223ffbf881a7721996f22dac518594e
|
3 |
+
size 1839
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-lsafy50k/sandbox/local_flytekit/d656dfcde256ff740a41645b0a35573b/00000
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
|
3 |
+
size 133886409
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-mkrrkkxt/sandbox/local_flytekit/d8175d827abaf1ea3da2c455259e80c6/00000
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
|
3 |
+
size 133886409
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-rpzbckua/sandbox/local_flytekit/6cf5cd5b55d6331ecbd20837130884b8/00000
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
|
3 |
+
size 133886409
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-so3w79m7/sandbox/local_flytekit/32f5ab2fae65d042c0a6d22025dea3ac/00000
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
|
3 |
+
size 133886409
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-udcr2wl6/sandbox/local_flytekit/1e3263ea13800320141857b26ef32688/00000
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
|
3 |
+
size 133886409
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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": 10, "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": ""}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/added_tokens.json
ADDED
File without changes
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/merges.txt
ADDED
File without changes
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/.no_exist/094fbdd0c911feb485ce55de1952ab2e75277e1e/vocab.json
ADDED
File without changes
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/0204ed10c186a4c7c68f55dff8f26087a45898d6.lock
ADDED
File without changes
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/08015a39463145a657e88a093aad79df9b2a76bf
ADDED
The diff for this file is too large to render.
See raw diff
|
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/08015a39463145a657e88a093aad79df9b2a76bf.lock
ADDED
File without changes
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/344b8478e86f0cb7933a50ef1fff9127ed27266c.lock
ADDED
File without changes
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/7cb72eb7357994a07d5c54ffdeef99dec84e21ae.lock
ADDED
File without changes
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b.lock
ADDED
File without changes
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/refs/main
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
094fbdd0c911feb485ce55de1952ab2e75277e1e
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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
|
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:eeaa01593806267e4358ec662b32448529e9521f7f15a6f131021450b6196cc2
|
3 |
+
size 5686106713
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/trainer_state.json
ADDED
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 6.666666666666667,
|
5 |
+
"global_step": 10,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 6.67,
|
12 |
+
"learning_rate": 2e-05,
|
13 |
+
"loss": 2.0319,
|
14 |
+
"step": 10
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 6.67,
|
18 |
+
"step": 10,
|
19 |
+
"total_flos": 4181427486720.0,
|
20 |
+
"train_loss": 2.031925582885742,
|
21 |
+
"train_runtime": 431.4804,
|
22 |
+
"train_samples_per_second": 4.38,
|
23 |
+
"train_steps_per_second": 0.023
|
24 |
+
}
|
25 |
+
],
|
26 |
+
"max_steps": 10,
|
27 |
+
"num_train_epochs": 10,
|
28 |
+
"total_flos": 4181427486720.0,
|
29 |
+
"trial_name": null,
|
30 |
+
"trial_params": null
|
31 |
+
}
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:54edde13ce02c45f23a8f5849df5058faeed6968da673a0576bf1bf3d8463d11
|
3 |
+
size 5563
|
flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/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":
|
|
|
1 |
+
{"base_model": "togethercomputer/RedPajama-INCITE-Base-3B-v1", "data_path": "wikipedia", "data_name": "20220301.simple", "num_epochs": 10, "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:
|
3 |
size 3169529757
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:181b39c52807ccfd435feeb305a7f40cf413de585b000c765c8a344efec01e8e
|
3 |
size 3169529757
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3899
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a490400de3d4575be6395dc868b65fefc72258ff5492e667d67f78f642804ea9
|
3 |
size 3899
|