Upload folder using huggingface_hub

.gitattributes

@@ -77,3 +77,12 @@ flyte7z9neqk5/local_flytekit/2002aa999dad15fe8dfd5901c73c9d2a/flyte-uhboju04/san
 flyte7z9neqk5/local_flytekit/2002aa999dad15fe8dfd5901c73c9d2a/flyte-uzrqm4fm/sandbox/local_flytekit/aa57f7d5c2606666a9c805e3219e4290/00000 filter=lfs diff=lfs merge=lfs -text
 flyte7z9neqk5/local_flytekit/2002aa999dad15fe8dfd5901c73c9d2a/flyte-y_wtkdev/sandbox/local_flytekit/4314274ef766fffe04a99df31a991f30/00000 filter=lfs diff=lfs merge=lfs -text
 flyte7z9neqk5/local_flytekit/2002aa999dad15fe8dfd5901c73c9d2a/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b filter=lfs diff=lfs merge=lfs -text
+flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-8ujlfsfb/sandbox/local_flytekit/be2e6207264225012cdb4050172b59f4/00000 filter=lfs diff=lfs merge=lfs -text
+flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-cw4395jo/sandbox/local_flytekit/d10d9d6c44a8c555d53fb2ea94cd4e81/00000 filter=lfs diff=lfs merge=lfs -text
+flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-ibnp8pcr/sandbox/local_flytekit/0998fa32059943ea2dccd958aea72125/00000 filter=lfs diff=lfs merge=lfs -text
+flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-kmwfk1ag/sandbox/local_flytekit/0e319c778bf4b3fe4a7cf129146b01e9/00000 filter=lfs diff=lfs merge=lfs -text
+flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-m4gfciv9/sandbox/local_flytekit/4ebcc6391c3fd174d4bc8e0c9b4f5477/00000 filter=lfs diff=lfs merge=lfs -text
+flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-n_5ajos6/sandbox/local_flytekit/2e8c3fd5e3770766775c6586d781c654/00000 filter=lfs diff=lfs merge=lfs -text
+flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-nfrsb6tj/sandbox/local_flytekit/ac8e394f8d5951aa902d8954089f2ee7/00000 filter=lfs diff=lfs merge=lfs -text
+flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-yfnwy3oa/sandbox/local_flytekit/92cdb3caaae47f47d0ec3ea5ea850d37/00000 filter=lfs diff=lfs merge=lfs -text
+flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/844708c42241de49775b872663ad58422a35b8e5a7d50b1f808408db8c1fb84b filter=lfs diff=lfs merge=lfs -text

flyte_training_config.json

@@ -1 +1 @@
-{"base_model": "togethercomputer/RedPajama-INCITE-Base-3B-v1", "data_path": "wikipedia", "data_name": "20220301.simple", "num_epochs": 1, "max_steps": 500, "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": 1, "max_steps": 100, "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": ""}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/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
+}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-8ujlfsfb/sandbox/local_flytekit/be2e6207264225012cdb4050172b59f4/00000

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-8v1m8uhd/sandbox/local_flytekit/inputs.pb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae981f0d56aca7c11d756b67e88f1066b061abb7b3f2d9926b89762e91524376
+size 1839

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-cw4395jo/sandbox/local_flytekit/d10d9d6c44a8c555d53fb2ea94cd4e81/00000

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-ibnp8pcr/sandbox/local_flytekit/0998fa32059943ea2dccd958aea72125/00000

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-kmwfk1ag/sandbox/local_flytekit/0e319c778bf4b3fe4a7cf129146b01e9/00000

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-m4gfciv9/sandbox/local_flytekit/4ebcc6391c3fd174d4bc8e0c9b4f5477/00000

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-n_5ajos6/sandbox/local_flytekit/2e8c3fd5e3770766775c6586d781c654/00000

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-nfrsb6tj/sandbox/local_flytekit/ac8e394f8d5951aa902d8954089f2ee7/00000

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte-yfnwy3oa/sandbox/local_flytekit/92cdb3caaae47f47d0ec3ea5ea850d37/00000

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/flyte_training_config.json

@@ -0,0 +1 @@
+{"base_model": "togethercomputer/RedPajama-INCITE-Base-3B-v1", "data_path": "wikipedia", "data_name": "20220301.simple", "num_epochs": 1, "max_steps": 100, "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": ""}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/generation_config.json

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/blobs/0204ed10c186a4c7c68f55dff8f26087a45898d6

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/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"
+}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/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
+}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/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|>"
+}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/models--togethercomputer--RedPajama-INCITE-Base-3B-v1/refs/main

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/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
+}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/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"
+}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/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|>"
+}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/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|>"
+}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e63cb70257a1b71bfddd7ac332cf76d74e1d295fe8c1bd2752e1b18938257b47
+size 5686106713

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/special_tokens_map.json

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

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/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|>"
+}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/trainer_state.json

@@ -0,0 +1,85 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 66.66666666666667,
+  "global_step": 100,
+  "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": 1.9669,
+      "step": 10
+    },
+    {
+      "epoch": 13.33,
+      "learning_rate": 2e-05,
+      "loss": 0.9447,
+      "step": 20
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 2e-05,
+      "loss": 0.1908,
+      "step": 30
+    },
+    {
+      "epoch": 26.67,
+      "learning_rate": 2e-05,
+      "loss": 0.0666,
+      "step": 40
+    },
+    {
+      "epoch": 33.33,
+      "learning_rate": 2e-05,
+      "loss": 0.0441,
+      "step": 50
+    },
+    {
+      "epoch": 40.0,
+      "learning_rate": 2e-05,
+      "loss": 0.0329,
+      "step": 60
+    },
+    {
+      "epoch": 46.67,
+      "learning_rate": 2e-05,
+      "loss": 0.0251,
+      "step": 70
+    },
+    {
+      "epoch": 53.33,
+      "learning_rate": 2e-05,
+      "loss": 0.019,
+      "step": 80
+    },
+    {
+      "epoch": 60.0,
+      "learning_rate": 2e-05,
+      "loss": 0.0166,
+      "step": 90
+    },
+    {
+      "epoch": 66.67,
+      "learning_rate": 2e-05,
+      "loss": 0.0133,
+      "step": 100
+    },
+    {
+      "epoch": 66.67,
+      "step": 100,
+      "total_flos": 42050959441920.0,
+      "train_loss": 0.3319866207242012,
+      "train_runtime": 4269.4137,
+      "train_samples_per_second": 2.998,
+      "train_steps_per_second": 0.023
+    }
+  ],
+  "max_steps": 100,
+  "num_train_epochs": 100,
+  "total_flos": 42050959441920.0,
+  "trial_name": null,
+  "trial_params": null
+}

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:64e0c2d603edfa479a939173d26bb532d940f3ac3258651395bfbf4755154a1a
+size 5499

flytet2ycrlv6/local_flytekit/faa0c0e4bee4bd512a5fe448c325d4ac/version.txt

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

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:581ac36bf92025df690933d8d3901195bf87753f05f362b567184032ced2fe64
+oid sha256:d8154c041b3d8a10f8ae5c009280f009c4f255f75efeb6bc7c938a4e9b2ca607
 size 3169529757

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:097b7a0e66a6317c1b276804d881ac420a8f9495acbd15a2e243540b20c6da0d
+oid sha256:fb6172aa47e693f345f3ab7e0de495c398686148ec579b81b54bacd169a7b67e
 size 3899