Upload folder using huggingface_hub

.gitattributes

@@ -59,3 +59,12 @@ flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-pbs7il4k/san
 flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-uruixz18/sandbox/local_flytekit/1525550a99b021eb4391c0302d22d55b/00000 filter=lfs diff=lfs merge=lfs -text
 flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/flyte-z4pk0owb/sandbox/local_flytekit/ff531e1dc85976e00f0b75473056a120/00000 filter=lfs diff=lfs merge=lfs -text
 flyte5spem3e8/local_flytekit/6ef5b51f616ae44965d7a8004c1606a6/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b filter=lfs diff=lfs merge=lfs -text
+flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-d3lu49g2/sandbox/local_flytekit/7202bf829fca4ce2b2d6f5eccbdca304/00000 filter=lfs diff=lfs merge=lfs -text
+flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-g0d6z68v/sandbox/local_flytekit/303f0bdbd1297f390becb36ed2b37a0a/00000 filter=lfs diff=lfs merge=lfs -text
+flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-g2fb47m8/sandbox/local_flytekit/e46090cfa982a6f7f7e471e7cc5452c7/00000 filter=lfs diff=lfs merge=lfs -text
+flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-lsafy50k/sandbox/local_flytekit/d656dfcde256ff740a41645b0a35573b/00000 filter=lfs diff=lfs merge=lfs -text
+flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-mkrrkkxt/sandbox/local_flytekit/d8175d827abaf1ea3da2c455259e80c6/00000 filter=lfs diff=lfs merge=lfs -text
+flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-rpzbckua/sandbox/local_flytekit/6cf5cd5b55d6331ecbd20837130884b8/00000 filter=lfs diff=lfs merge=lfs -text
+flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-so3w79m7/sandbox/local_flytekit/32f5ab2fae65d042c0a6d22025dea3ac/00000 filter=lfs diff=lfs merge=lfs -text
+flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-udcr2wl6/sandbox/local_flytekit/1e3263ea13800320141857b26ef32688/00000 filter=lfs diff=lfs merge=lfs -text
+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

@@ -0,0 +1,26 @@
+{
+  "_name_or_path": "togethercomputer/RedPajama-INCITE-Base-3B-v1",
+  "architectures": [
+    "GPTNeoXForCausalLM"
+  ],
+  "bos_token_id": 0,
+  "classifier_dropout": 0.1,
+  "eos_token_id": 0,
+  "hidden_act": "gelu",
+  "hidden_size": 2560,
+  "initializer_range": 0.02,
+  "intermediate_size": 10240,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 2048,
+  "model_type": "gpt_neox",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "rotary_emb_base": 10000,
+  "rotary_pct": 1.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.29.2",
+  "use_cache": true,
+  "use_parallel_residual": false,
+  "vocab_size": 50432
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-d3lu49g2/sandbox/local_flytekit/7202bf829fca4ce2b2d6f5eccbdca304/00000

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
+size 133886409

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-g0d6z68v/sandbox/local_flytekit/303f0bdbd1297f390becb36ed2b37a0a/00000

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
+size 133886409

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-g2fb47m8/sandbox/local_flytekit/e46090cfa982a6f7f7e471e7cc5452c7/00000

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
+size 133886409

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-kpzvlk9q/sandbox/local_flytekit/inputs.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:db8a93ad392ab06c25eac142df7874c95223ffbf881a7721996f22dac518594e
+size 1839

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-lsafy50k/sandbox/local_flytekit/d656dfcde256ff740a41645b0a35573b/00000

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
+size 133886409

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-mkrrkkxt/sandbox/local_flytekit/d8175d827abaf1ea3da2c455259e80c6/00000

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
+size 133886409

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-rpzbckua/sandbox/local_flytekit/6cf5cd5b55d6331ecbd20837130884b8/00000

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
+size 133886409

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-so3w79m7/sandbox/local_flytekit/32f5ab2fae65d042c0a6d22025dea3ac/00000

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
+size 133886409

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte-udcr2wl6/sandbox/local_flytekit/1e3263ea13800320141857b26ef32688/00000

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:067772915d011157436dc1ea88cb38756555e25be2d07616d1ee97dfac6e6535
+size 133886409

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/flyte_training_config.json

@@ -0,0 +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

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 0,
+  "eos_token_id": 0,
+  "transformers_version": "4.29.2"
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/0204ed10c186a4c7c68f55dff8f26087a45898d6

@@ -0,0 +1,5 @@
+{
+  "bos_token": "<|endoftext|>",
+  "eos_token": "<|endoftext|>",
+  "unk_token": "<|endoftext|>"
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/344b8478e86f0cb7933a50ef1fff9127ed27266c

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 0,
+  "eos_token_id": 0,
+  "transformers_version": "4.28.1"
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/7cb72eb7357994a07d5c54ffdeef99dec84e21ae

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "rp_3b_800b",
+  "architectures": [
+    "GPTNeoXForCausalLM"
+  ],
+  "bos_token_id": 0,
+  "eos_token_id": 0,
+  "hidden_act": "gelu",
+  "hidden_size": 2560,
+  "initializer_range": 0.02,
+  "intermediate_size": 10240,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 2048,
+  "model_type": "gpt_neox",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "rotary_emb_base": 10000,
+  "rotary_pct": 1.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.28.1",
+  "use_cache": true,
+  "use_parallel_residual": false,
+  "vocab_size": 50432
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b
+size 5686113497

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/af28df4f9190240e1c3fb77a27b6f3d9bdd41c2d

@@ -0,0 +1,9 @@
+{
+  "add_prefix_space": false,
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "model_max_length": 2048,
+  "tokenizer_class": "GPTNeoXTokenizer",
+  "unk_token": "<|endoftext|>"
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/refs/main

@@ -0,0 +1 @@
+094fbdd0c911feb485ce55de1952ab2e75277e1e

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "rp_3b_800b",
+  "architectures": [
+    "GPTNeoXForCausalLM"
+  ],
+  "bos_token_id": 0,
+  "eos_token_id": 0,
+  "hidden_act": "gelu",
+  "hidden_size": 2560,
+  "initializer_range": 0.02,
+  "intermediate_size": 10240,
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 2048,
+  "model_type": "gpt_neox",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "rotary_emb_base": 10000,
+  "rotary_pct": 1.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.28.1",
+  "use_cache": true,
+  "use_parallel_residual": false,
+  "vocab_size": 50432
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/generation_config.json

@@ -0,0 +1,6 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 0,
+  "eos_token_id": 0,
+  "transformers_version": "4.28.1"
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b
+size 5686113497

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/special_tokens_map.json

@@ -0,0 +1,5 @@
+{
+  "bos_token": "<|endoftext|>",
+  "eos_token": "<|endoftext|>",
+  "unk_token": "<|endoftext|>"
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/snapshots/094fbdd0c911feb485ce55de1952ab2e75277e1e/tokenizer_config.json

@@ -0,0 +1,9 @@
+{
+  "add_prefix_space": false,
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "model_max_length": 2048,
+  "tokenizer_class": "GPTNeoXTokenizer",
+  "unk_token": "<|endoftext|>"
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eeaa01593806267e4358ec662b32448529e9521f7f15a6f131021450b6196cc2
+size 5686106713

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/special_tokens_map.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": "<|endoftext|>",
+  "eos_token": "<|endoftext|>",
+  "pad_token": "[PAD]",
+  "unk_token": "<|endoftext|>"
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/tokenizer_config.json

@@ -0,0 +1,11 @@
+{
+  "add_prefix_space": false,
+  "bos_token": "<|endoftext|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|endoftext|>",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "padding_side": "right",
+  "tokenizer_class": "GPTNeoXTokenizer",
+  "unk_token": "<|endoftext|>"
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/trainer_state.json

@@ -0,0 +1,31 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 6.666666666666667,
+  "global_step": 10,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 6.67,
+      "learning_rate": 2e-05,
+      "loss": 2.0319,
+      "step": 10
+    },
+    {
+      "epoch": 6.67,
+      "step": 10,
+      "total_flos": 4181427486720.0,
+      "train_loss": 2.031925582885742,
+      "train_runtime": 431.4804,
+      "train_samples_per_second": 4.38,
+      "train_steps_per_second": 0.023
+    }
+  ],
+  "max_steps": 10,
+  "num_train_epochs": 10,
+  "total_flos": 4181427486720.0,
+  "trial_name": null,
+  "trial_params": null
+}

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54edde13ce02c45f23a8f5849df5058faeed6968da673a0576bf1bf3d8463d11
+size 5563

flyte7xlnw_t7/local_flytekit/4ab756a0f57dd11dd93940145c628b70/version.txt

@@ -0,0 +1 @@
+1

flyte_training_config.json

@@ -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": ""}
+{"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

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4c7fed73f3ca052abf4965428545bffca4951d6cdd4d7bc1316e4eedb7d08637
+oid sha256:181b39c52807ccfd435feeb305a7f40cf413de585b000c765c8a344efec01e8e
 size 3169529757

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:08b08062463650c20488fac34e4e4253f42e5221a2d85fb32be0717ceb253039
+oid sha256:a490400de3d4575be6395dc868b65fefc72258ff5492e667d67f78f642804ea9
 size 3899