Initial Commit Model

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

config.json

@@ -0,0 +1,39 @@
+{
+  "_name_or_path": "configs/sled_mt5_config.json",
+  "architectures": [
+    "SledForConditionalGeneration"
+  ],
+  "context_size": 256,
+  "d_ff": 2048,
+  "d_kv": 64,
+  "d_model": 768,
+  "decoder_start_token_id": 0,
+  "dense_act_fn": "gelu_new",
+  "dropout_rate": 0.1,
+  "encode_prefix": true,
+  "eos_token_id": 1,
+  "feed_forward_proj": "gated-gelu",
+  "initializer_factor": 1.0,
+  "is_encoder_decoder": true,
+  "is_gated_act": true,
+  "layer_norm_epsilon": 1e-06,
+  "max_target_length": 2048,
+  "model_type": "tau/sled",
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "prepend_prefix": true,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "sliding_method": "dynamic",
+  "tie_word_embeddings": false,
+  "tokenizer_class": "T5Tokenizer",
+  "torch_dtype": "float32",
+  "transformers_version": "4.28.1",
+  "underlying_config": "google/mt5-base",
+  "use_cache": false,
+  "vocab_size": 250100,
+  "window_fraction": 0.5
+}

generation_config.json

@@ -0,0 +1,8 @@
+{
+  "_from_model_config": true,
+  "decoder_start_token_id": 0,
+  "eos_token_id": 1,
+  "pad_token_id": 0,
+  "transformers_version": "4.28.1",
+  "use_cache": false
+}

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:063cb29821aecc9a67b48d4d0b14349e0a3a8c1f28eea8e0962ee5f6e07878db
+size 1962672128

pytorch_model-00001-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e662394d45079d7e582bab2a641664eeda72d54ee8276f112703612cddac080d
+size 1961099264

pytorch_model-00002-of-00002.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9744048321e91cb43c39270ae6c303e336f337d5ffa87235b785c93bd686af05
+size 1961361408

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c64a9cfbe7f682b13a547eb60fd69fece959ceade11960d8fcb8ded3a6b39924
+size 14575

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eab0c8af771dd5211bdac45c5a7027b1a54ae9fd05519b4acf3e216d3b8d0341
+size 627

special_tokens_map.json

@@ -0,0 +1,5 @@
+{
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
+}

spiece.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef78f86560d809067d12bac6c09f19a462cb3af3f54d2b8acbba26e1433125d6
+size 4309802

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93c3578052e1605d8332eb961bc08d72e246071974e4cc54aa6991826b802aa5
+size 16330369

tokenizer_config.json

@@ -0,0 +1,11 @@
+{
+  "additional_special_tokens": null,
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "</s>",
+  "extra_ids": 0,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<pad>",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "T5Tokenizer",
+  "unk_token": "<unk>"
+}

trainer_state.json

@@ -0,0 +1,229 @@
+{
+  "best_metric": 28.7934,
+  "best_model_checkpoint": "models/sled/output_sled_mt5_base2/checkpoint-11985",
+  "epoch": 5.0,
+  "global_step": 11985,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.3904338153503894e-05,
+      "loss": 18.6448,
+      "step": 500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.7808676307007788e-05,
+      "loss": 3.8147,
+      "step": 1000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.171301446051168e-05,
+      "loss": 2.6096,
+      "step": 1500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.5617352614015575e-05,
+      "loss": 2.3758,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0,
+      "eval_mean_prediction_length_characters": 7800.0,
+      "eval_mean_prediction_length_tokens": 2047.0,
+      "eval_num_predicted": 1,
+      "eval_rouge/geometric_mean": 14.0723,
+      "eval_rouge/rouge1": 24.3546,
+      "eval_rouge/rouge2": 7.1185,
+      "eval_rouge/rougeL": 16.074,
+      "eval_rouge/rougeLsum": 21.4321,
+      "eval_runtime": 358.1416,
+      "eval_samples_per_second": 0.003,
+      "eval_steps_per_second": 0.003,
+      "step": 2397
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 6.952169076751947e-05,
+      "loss": 2.1056,
+      "step": 2500
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 8.342602892102336e-05,
+      "loss": 2.0678,
+      "step": 3000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.733036707452726e-05,
+      "loss": 1.9603,
+      "step": 3500
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.87515065360487e-05,
+      "loss": 1.8989,
+      "step": 4000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.720634135789116e-05,
+      "loss": 1.8726,
+      "step": 4500
+    },
+    {
+      "epoch": 2.0,
+      "eval_mean_prediction_length_characters": 7920.0,
+      "eval_mean_prediction_length_tokens": 2047.0,
+      "eval_num_predicted": 1,
+      "eval_rouge/geometric_mean": 9.5424,
+      "eval_rouge/rouge1": 14.5332,
+      "eval_rouge/rouge2": 5.973,
+      "eval_rouge/rougeL": 10.0096,
+      "eval_rouge/rougeLsum": 13.9557,
+      "eval_runtime": 361.8939,
+      "eval_samples_per_second": 0.003,
+      "eval_steps_per_second": 0.003,
+      "step": 4794
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 9.566117617973361e-05,
+      "loss": 1.7568,
+      "step": 5000
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 9.411601100157608e-05,
+      "loss": 1.7487,
+      "step": 5500
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 9.257084582341852e-05,
+      "loss": 1.7171,
+      "step": 6000
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 9.102568064526098e-05,
+      "loss": 1.7211,
+      "step": 6500
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 8.948051546710344e-05,
+      "loss": 1.6807,
+      "step": 7000
+    },
+    {
+      "epoch": 3.0,
+      "eval_mean_prediction_length_characters": 7675.0,
+      "eval_mean_prediction_length_tokens": 2047.0,
+      "eval_num_predicted": 1,
+      "eval_rouge/geometric_mean": 13.6689,
+      "eval_rouge/rouge1": 24.0,
+      "eval_rouge/rouge2": 7.1289,
+      "eval_rouge/rougeL": 14.9268,
+      "eval_rouge/rougeLsum": 21.2683,
+      "eval_runtime": 361.2079,
+      "eval_samples_per_second": 0.003,
+      "eval_steps_per_second": 0.003,
+      "step": 7191
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 8.79353502889459e-05,
+      "loss": 1.6141,
+      "step": 7500
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 8.639018511078834e-05,
+      "loss": 1.6026,
+      "step": 8000
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 8.48450199326308e-05,
+      "loss": 1.616,
+      "step": 8500
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 8.329985475447326e-05,
+      "loss": 1.6012,
+      "step": 9000
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 8.175468957631572e-05,
+      "loss": 1.6069,
+      "step": 9500
+    },
+    {
+      "epoch": 4.0,
+      "eval_mean_prediction_length_characters": 9281.0,
+      "eval_mean_prediction_length_tokens": 2047.0,
+      "eval_num_predicted": 1,
+      "eval_rouge/geometric_mean": 16.6758,
+      "eval_rouge/rouge1": 26.9307,
+      "eval_rouge/rouge2": 10.6046,
+      "eval_rouge/rougeL": 16.2376,
+      "eval_rouge/rougeLsum": 24.5545,
+      "eval_runtime": 363.6434,
+      "eval_samples_per_second": 0.003,
+      "eval_steps_per_second": 0.003,
+      "step": 9588
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.020952439815816e-05,
+      "loss": 1.5703,
+      "step": 10000
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 7.866435922000062e-05,
+      "loss": 1.5467,
+      "step": 10500
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 7.711919404184308e-05,
+      "loss": 1.5567,
+      "step": 11000
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 7.557402886368554e-05,
+      "loss": 1.5021,
+      "step": 11500
+    },
+    {
+      "epoch": 5.0,
+      "eval_mean_prediction_length_characters": 7806.0,
+      "eval_mean_prediction_length_tokens": 2047.0,
+      "eval_num_predicted": 1,
+      "eval_rouge/geometric_mean": 28.7934,
+      "eval_rouge/rouge1": 48.5567,
+      "eval_rouge/rouge2": 21.775,
+      "eval_rouge/rougeL": 22.5773,
+      "eval_rouge/rougeLsum": 46.1856,
+      "eval_runtime": 395.7669,
+      "eval_samples_per_second": 0.003,
+      "eval_steps_per_second": 0.003,
+      "step": 11985
+    }
+  ],
+  "max_steps": 35955,
+  "num_train_epochs": 15,
+  "total_flos": 4.0365738632318976e+17,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf08ccca604f1b6e5490b1bb46a942c367110f2fb76adc88496f87f873d6f525
+size 3899