add updated checkpoints

.gitattributes

@@ -33,12 +33,8 @@ 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
-checkpoint-5000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
-checkpoint-5000/rng_state.pth filter=lfs diff=lfs merge=lfs -text
-checkpoint-5000/scheduler.pt filter=lfs diff=lfs merge=lfs -text
-checkpoint-5000/trainer_state.json filter=lfs diff=lfs merge=lfs -text
-checkpoint-5000/training_args.bin filter=lfs diff=lfs merge=lfs -text
-checkpoint-5000/config.json filter=lfs diff=lfs merge=lfs -text
-checkpoint-5000/generation_config.json filter=lfs diff=lfs merge=lfs -text
-checkpoint-5000/optimizer.pt filter=lfs diff=lfs merge=lfs -text
 pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+checkpoint-5000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+old/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+checkpoint-10000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
+checkpoint-15000/pytorch_model.bin filter=lfs diff=lfs merge=lfs -text

checkpoint-10000/config.json

@@ -0,0 +1,62 @@
+{
+  "_name_or_path": "google/flan-t5-base",
+  "architectures": [
+    "T5ForConditionalGeneration"
+  ],
+  "classifier_dropout": 0.0,
+  "d_ff": 2048,
+  "d_kv": 64,
+  "d_model": 768,
+  "decoder_start_token_id": 0,
+  "dense_act_fn": "gelu_new",
+  "dropout_rate": 0.1,
+  "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,
+  "model_type": "t5",
+  "n_positions": 512,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "task_specific_params": {
+    "summarization": {
+      "early_stopping": true,
+      "length_penalty": 2.0,
+      "max_length": 200,
+      "min_length": 30,
+      "no_repeat_ngram_size": 3,
+      "num_beams": 4,
+      "prefix": "summarize: "
+    },
+    "translation_en_to_de": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to German: "
+    },
+    "translation_en_to_fr": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to French: "
+    },
+    "translation_en_to_ro": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to Romanian: "
+    }
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.33.1",
+  "use_cache": true,
+  "vocab_size": 32128
+}

checkpoint-10000/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "decoder_start_token_id": 0,
+  "eos_token_id": 1,
+  "pad_token_id": 0,
+  "transformers_version": "4.33.1"
+}

checkpoint-10000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93e3449b196ded472f0c110cfb9c227458c438e4af5d69bf6df86b14537514ae
+size 2371333

checkpoint-10000/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:02140318cb07e579401ae4518cddbaac6194ec5970d26c2862b50f1668d35431
+size 990408885

checkpoint-10000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7a55edb6d94fca3aa6bf22b9c2bef466f4db503e6ce7363460dc43b39fae5b0
+size 14575

checkpoint-10000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ca84735770f2a357642bb07a692869b77f0409ed01b19bf3c890697e28d947e
+size 627

checkpoint-10000/trainer_state.json

@@ -0,0 +1,299 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.2597631645250693,
+  "eval_steps": 500,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000975,
+      "loss": 1.0238,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.8929525017738342,
+      "eval_runtime": 4.0089,
+      "eval_samples_per_second": 74.833,
+      "eval_steps_per_second": 37.417,
+      "step": 500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00095,
+      "loss": 0.9971,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.8510345220565796,
+      "eval_runtime": 4.0089,
+      "eval_samples_per_second": 74.833,
+      "eval_steps_per_second": 37.417,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.000925,
+      "loss": 0.9522,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.8565580248832703,
+      "eval_runtime": 4.0278,
+      "eval_samples_per_second": 74.483,
+      "eval_steps_per_second": 37.241,
+      "step": 1500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0009000000000000001,
+      "loss": 0.9143,
+      "step": 2000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.8810315132141113,
+      "eval_runtime": 4.0097,
+      "eval_samples_per_second": 74.819,
+      "eval_steps_per_second": 37.409,
+      "step": 2000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000875,
+      "loss": 0.9151,
+      "step": 2500
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.8461073040962219,
+      "eval_runtime": 4.0108,
+      "eval_samples_per_second": 74.798,
+      "eval_steps_per_second": 37.399,
+      "step": 2500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00085,
+      "loss": 0.9263,
+      "step": 3000
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.8521466255187988,
+      "eval_runtime": 4.0059,
+      "eval_samples_per_second": 74.89,
+      "eval_steps_per_second": 37.445,
+      "step": 3000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000825,
+      "loss": 0.9102,
+      "step": 3500
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.8521981835365295,
+      "eval_runtime": 4.0413,
+      "eval_samples_per_second": 74.234,
+      "eval_steps_per_second": 37.117,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0008,
+      "loss": 0.8866,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.8259674906730652,
+      "eval_runtime": 4.0316,
+      "eval_samples_per_second": 74.412,
+      "eval_steps_per_second": 37.206,
+      "step": 4000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0007750000000000001,
+      "loss": 0.911,
+      "step": 4500
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.8008900284767151,
+      "eval_runtime": 4.0224,
+      "eval_samples_per_second": 74.583,
+      "eval_steps_per_second": 37.292,
+      "step": 4500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00075,
+      "loss": 0.8536,
+      "step": 5000
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.808663547039032,
+      "eval_runtime": 5.0615,
+      "eval_samples_per_second": 59.271,
+      "eval_steps_per_second": 29.635,
+      "step": 5000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.000725,
+      "loss": 0.8277,
+      "step": 5500
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.7903586626052856,
+      "eval_runtime": 4.0081,
+      "eval_samples_per_second": 74.849,
+      "eval_steps_per_second": 37.425,
+      "step": 5500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0007,
+      "loss": 0.8413,
+      "step": 6000
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.8013813495635986,
+      "eval_runtime": 3.994,
+      "eval_samples_per_second": 75.113,
+      "eval_steps_per_second": 37.556,
+      "step": 6000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.000675,
+      "loss": 0.8491,
+      "step": 6500
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 0.7867017984390259,
+      "eval_runtime": 4.0072,
+      "eval_samples_per_second": 74.864,
+      "eval_steps_per_second": 37.432,
+      "step": 6500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0006500000000000001,
+      "loss": 0.8077,
+      "step": 7000
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.759165346622467,
+      "eval_runtime": 4.0166,
+      "eval_samples_per_second": 74.691,
+      "eval_steps_per_second": 37.345,
+      "step": 7000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.000625,
+      "loss": 0.796,
+      "step": 7500
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.7527943849563599,
+      "eval_runtime": 3.998,
+      "eval_samples_per_second": 75.037,
+      "eval_steps_per_second": 37.518,
+      "step": 7500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0006,
+      "loss": 0.7952,
+      "step": 8000
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 0.7407393455505371,
+      "eval_runtime": 3.9421,
+      "eval_samples_per_second": 76.102,
+      "eval_steps_per_second": 38.051,
+      "step": 8000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.000575,
+      "loss": 0.5761,
+      "step": 8500
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 0.7574229836463928,
+      "eval_runtime": 3.9855,
+      "eval_samples_per_second": 75.273,
+      "eval_steps_per_second": 37.636,
+      "step": 8500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00055,
+      "loss": 0.6084,
+      "step": 9000
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 0.7349967956542969,
+      "eval_runtime": 3.995,
+      "eval_samples_per_second": 75.093,
+      "eval_steps_per_second": 37.547,
+      "step": 9000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0005250000000000001,
+      "loss": 0.5863,
+      "step": 9500
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 0.7407773733139038,
+      "eval_runtime": 3.9874,
+      "eval_samples_per_second": 75.238,
+      "eval_steps_per_second": 37.619,
+      "step": 9500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0005,
+      "loss": 0.5835,
+      "step": 10000
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 0.719527006149292,
+      "eval_runtime": 3.9471,
+      "eval_samples_per_second": 76.006,
+      "eval_steps_per_second": 38.003,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 20000,
+  "num_train_epochs": 3,
+  "save_steps": 5000,
+  "total_flos": 1.773098919161856e+16,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-10000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a0a0ea9dbd2491eec13fb387c65610f8eb3edd4ca287eecb3bede1dd0d34916
+size 4155

checkpoint-15000/config.json

@@ -0,0 +1,62 @@
+{
+  "_name_or_path": "google/flan-t5-base",
+  "architectures": [
+    "T5ForConditionalGeneration"
+  ],
+  "classifier_dropout": 0.0,
+  "d_ff": 2048,
+  "d_kv": 64,
+  "d_model": 768,
+  "decoder_start_token_id": 0,
+  "dense_act_fn": "gelu_new",
+  "dropout_rate": 0.1,
+  "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,
+  "model_type": "t5",
+  "n_positions": 512,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "task_specific_params": {
+    "summarization": {
+      "early_stopping": true,
+      "length_penalty": 2.0,
+      "max_length": 200,
+      "min_length": 30,
+      "no_repeat_ngram_size": 3,
+      "num_beams": 4,
+      "prefix": "summarize: "
+    },
+    "translation_en_to_de": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to German: "
+    },
+    "translation_en_to_fr": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to French: "
+    },
+    "translation_en_to_ro": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to Romanian: "
+    }
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.33.1",
+  "use_cache": true,
+  "vocab_size": 32128
+}

checkpoint-15000/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "decoder_start_token_id": 0,
+  "eos_token_id": 1,
+  "pad_token_id": 0,
+  "transformers_version": "4.33.1"
+}

checkpoint-15000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e164f9fad0641e64a32b9d367cf6e92483eb5cd7df7a4dd42c3ddddc0cadebe1
+size 2371333

checkpoint-15000/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:42f1dac77bf4d254203f8f6dd684ada1cf998f26b81addd2e8dd06b2eeab8cd6
+size 990408885

checkpoint-15000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fca6bd4f2027e9f6a64120d4dd9cfacab4778f895ae586fd5c13f7cff62aac59
+size 14575

checkpoint-15000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62b793f9ea6eba39185d82063b1e7434411e2aeca1bab5a010024f955d1696b3
+size 627

checkpoint-15000/trainer_state.json

@@ -0,0 +1,439 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.8896447467876039,
+  "eval_steps": 500,
+  "global_step": 15000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000975,
+      "loss": 1.0238,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.8929525017738342,
+      "eval_runtime": 4.0089,
+      "eval_samples_per_second": 74.833,
+      "eval_steps_per_second": 37.417,
+      "step": 500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00095,
+      "loss": 0.9971,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.8510345220565796,
+      "eval_runtime": 4.0089,
+      "eval_samples_per_second": 74.833,
+      "eval_steps_per_second": 37.417,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.000925,
+      "loss": 0.9522,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.8565580248832703,
+      "eval_runtime": 4.0278,
+      "eval_samples_per_second": 74.483,
+      "eval_steps_per_second": 37.241,
+      "step": 1500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0009000000000000001,
+      "loss": 0.9143,
+      "step": 2000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.8810315132141113,
+      "eval_runtime": 4.0097,
+      "eval_samples_per_second": 74.819,
+      "eval_steps_per_second": 37.409,
+      "step": 2000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000875,
+      "loss": 0.9151,
+      "step": 2500
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.8461073040962219,
+      "eval_runtime": 4.0108,
+      "eval_samples_per_second": 74.798,
+      "eval_steps_per_second": 37.399,
+      "step": 2500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00085,
+      "loss": 0.9263,
+      "step": 3000
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.8521466255187988,
+      "eval_runtime": 4.0059,
+      "eval_samples_per_second": 74.89,
+      "eval_steps_per_second": 37.445,
+      "step": 3000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000825,
+      "loss": 0.9102,
+      "step": 3500
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.8521981835365295,
+      "eval_runtime": 4.0413,
+      "eval_samples_per_second": 74.234,
+      "eval_steps_per_second": 37.117,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0008,
+      "loss": 0.8866,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.8259674906730652,
+      "eval_runtime": 4.0316,
+      "eval_samples_per_second": 74.412,
+      "eval_steps_per_second": 37.206,
+      "step": 4000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0007750000000000001,
+      "loss": 0.911,
+      "step": 4500
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.8008900284767151,
+      "eval_runtime": 4.0224,
+      "eval_samples_per_second": 74.583,
+      "eval_steps_per_second": 37.292,
+      "step": 4500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00075,
+      "loss": 0.8536,
+      "step": 5000
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.808663547039032,
+      "eval_runtime": 5.0615,
+      "eval_samples_per_second": 59.271,
+      "eval_steps_per_second": 29.635,
+      "step": 5000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.000725,
+      "loss": 0.8277,
+      "step": 5500
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.7903586626052856,
+      "eval_runtime": 4.0081,
+      "eval_samples_per_second": 74.849,
+      "eval_steps_per_second": 37.425,
+      "step": 5500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0007,
+      "loss": 0.8413,
+      "step": 6000
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.8013813495635986,
+      "eval_runtime": 3.994,
+      "eval_samples_per_second": 75.113,
+      "eval_steps_per_second": 37.556,
+      "step": 6000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.000675,
+      "loss": 0.8491,
+      "step": 6500
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 0.7867017984390259,
+      "eval_runtime": 4.0072,
+      "eval_samples_per_second": 74.864,
+      "eval_steps_per_second": 37.432,
+      "step": 6500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0006500000000000001,
+      "loss": 0.8077,
+      "step": 7000
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.759165346622467,
+      "eval_runtime": 4.0166,
+      "eval_samples_per_second": 74.691,
+      "eval_steps_per_second": 37.345,
+      "step": 7000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.000625,
+      "loss": 0.796,
+      "step": 7500
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.7527943849563599,
+      "eval_runtime": 3.998,
+      "eval_samples_per_second": 75.037,
+      "eval_steps_per_second": 37.518,
+      "step": 7500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0006,
+      "loss": 0.7952,
+      "step": 8000
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 0.7407393455505371,
+      "eval_runtime": 3.9421,
+      "eval_samples_per_second": 76.102,
+      "eval_steps_per_second": 38.051,
+      "step": 8000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.000575,
+      "loss": 0.5761,
+      "step": 8500
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 0.7574229836463928,
+      "eval_runtime": 3.9855,
+      "eval_samples_per_second": 75.273,
+      "eval_steps_per_second": 37.636,
+      "step": 8500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00055,
+      "loss": 0.6084,
+      "step": 9000
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 0.7349967956542969,
+      "eval_runtime": 3.995,
+      "eval_samples_per_second": 75.093,
+      "eval_steps_per_second": 37.547,
+      "step": 9000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0005250000000000001,
+      "loss": 0.5863,
+      "step": 9500
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 0.7407773733139038,
+      "eval_runtime": 3.9874,
+      "eval_samples_per_second": 75.238,
+      "eval_steps_per_second": 37.619,
+      "step": 9500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0005,
+      "loss": 0.5835,
+      "step": 10000
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 0.719527006149292,
+      "eval_runtime": 3.9471,
+      "eval_samples_per_second": 76.006,
+      "eval_steps_per_second": 38.003,
+      "step": 10000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.000475,
+      "loss": 0.5751,
+      "step": 10500
+    },
+    {
+      "epoch": 1.32,
+      "eval_loss": 0.7423205971717834,
+      "eval_runtime": 3.9978,
+      "eval_samples_per_second": 75.041,
+      "eval_steps_per_second": 37.52,
+      "step": 10500
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00045000000000000004,
+      "loss": 0.5746,
+      "step": 11000
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 0.7284040451049805,
+      "eval_runtime": 3.9657,
+      "eval_samples_per_second": 75.649,
+      "eval_steps_per_second": 37.824,
+      "step": 11000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.000425,
+      "loss": 0.5847,
+      "step": 11500
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 0.7247716188430786,
+      "eval_runtime": 3.962,
+      "eval_samples_per_second": 75.719,
+      "eval_steps_per_second": 37.86,
+      "step": 11500
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0004,
+      "loss": 0.5759,
+      "step": 12000
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 0.7151039838790894,
+      "eval_runtime": 3.9922,
+      "eval_samples_per_second": 75.147,
+      "eval_steps_per_second": 37.573,
+      "step": 12000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.000375,
+      "loss": 0.5922,
+      "step": 12500
+    },
+    {
+      "epoch": 1.57,
+      "eval_loss": 0.7031030058860779,
+      "eval_runtime": 4.0386,
+      "eval_samples_per_second": 74.282,
+      "eval_steps_per_second": 37.141,
+      "step": 12500
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00035,
+      "loss": 0.5678,
+      "step": 13000
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 0.710035502910614,
+      "eval_runtime": 4.0808,
+      "eval_samples_per_second": 73.515,
+      "eval_steps_per_second": 36.758,
+      "step": 13000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00032500000000000004,
+      "loss": 0.5627,
+      "step": 13500
+    },
+    {
+      "epoch": 1.7,
+      "eval_loss": 0.7093074321746826,
+      "eval_runtime": 4.0326,
+      "eval_samples_per_second": 74.393,
+      "eval_steps_per_second": 37.197,
+      "step": 13500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0003,
+      "loss": 0.5568,
+      "step": 14000
+    },
+    {
+      "epoch": 1.76,
+      "eval_loss": 0.6942310333251953,
+      "eval_runtime": 4.0875,
+      "eval_samples_per_second": 73.395,
+      "eval_steps_per_second": 36.697,
+      "step": 14000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.000275,
+      "loss": 0.5589,
+      "step": 14500
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 0.6947352290153503,
+      "eval_runtime": 3.9856,
+      "eval_samples_per_second": 75.27,
+      "eval_steps_per_second": 37.635,
+      "step": 14500
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00025,
+      "loss": 0.5675,
+      "step": 15000
+    },
+    {
+      "epoch": 1.89,
+      "eval_loss": 0.6902616620063782,
+      "eval_runtime": 4.0066,
+      "eval_samples_per_second": 74.876,
+      "eval_steps_per_second": 37.438,
+      "step": 15000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 20000,
+  "num_train_epochs": 3,
+  "save_steps": 5000,
+  "total_flos": 2.66437480937472e+16,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-15000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a0a0ea9dbd2491eec13fb387c65610f8eb3edd4ca287eecb3bede1dd0d34916
+size 4155

checkpoint-5000/config.json

@@ -1,3 +1,62 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5a816a18dcfe99ecf1c5ab86e22b3354b846cb2fdddc15329b6d7b0ce509de9b
-size 1528
+{
+  "_name_or_path": "google/flan-t5-base",
+  "architectures": [
+    "T5ForConditionalGeneration"
+  ],
+  "classifier_dropout": 0.0,
+  "d_ff": 2048,
+  "d_kv": 64,
+  "d_model": 768,
+  "decoder_start_token_id": 0,
+  "dense_act_fn": "gelu_new",
+  "dropout_rate": 0.1,
+  "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,
+  "model_type": "t5",
+  "n_positions": 512,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "task_specific_params": {
+    "summarization": {
+      "early_stopping": true,
+      "length_penalty": 2.0,
+      "max_length": 200,
+      "min_length": 30,
+      "no_repeat_ngram_size": 3,
+      "num_beams": 4,
+      "prefix": "summarize: "
+    },
+    "translation_en_to_de": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to German: "
+    },
+    "translation_en_to_fr": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to French: "
+    },
+    "translation_en_to_ro": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to Romanian: "
+    }
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.33.1",
+  "use_cache": true,
+  "vocab_size": 32128
+}

checkpoint-5000/generation_config.json

@@ -1,3 +1,7 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8f6833851ce53496a43ef87a975c766f7a3049e2d598ecef609a526ca6308534
-size 142
+{
+  "_from_model_config": true,
+  "decoder_start_token_id": 0,
+  "eos_token_id": 1,
+  "pad_token_id": 0,
+  "transformers_version": "4.33.1"
+}

checkpoint-5000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a98bad1336f5b5bd7166561dd9169ef0ae0458d84a12fe4495dafdb4c698c8e3
+oid sha256:f2d36b3a4675e30c3861b0b833b0ee1838830d4bbfcf92d69d39b4bba7c06442
 size 2371333

checkpoint-5000/pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7c8d2b4fc474f2a3ee4e727ef3beeb48bda1580929f0533595549493a670dd58
+oid sha256:efc5a5d32b2177423dcd6a5abb9e638dbfe38841c5c85b4c70a4a05c1a1ae9d0
 size 990408885

checkpoint-5000/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ad1e41a24146cf5c2e027973b4674ca7056995fc270e892208cda116dacb7e4f
+oid sha256:f114eb1bff9d2da7d00f2830d9792a82cd85c73cbe79c7fce753611d24277ca2
 size 14575

checkpoint-5000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:06319e8fb637c8a375199a70338dd58362987a6132a8f8ca31b00c3b0b013d9b
+oid sha256:cf2e868687b4ae6ba785ff2503f15cde78e0501a53544557c44b05f9fae02479
 size 627

checkpoint-5000/trainer_state.json

@@ -1,3 +1,159 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0896850485757f9eb02938e3a809abe3398bb4516d6af6b771aeffaf3094ae84
-size 3475
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6298815822625347,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000975,
+      "loss": 1.0238,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.8929525017738342,
+      "eval_runtime": 4.0089,
+      "eval_samples_per_second": 74.833,
+      "eval_steps_per_second": 37.417,
+      "step": 500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00095,
+      "loss": 0.9971,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.8510345220565796,
+      "eval_runtime": 4.0089,
+      "eval_samples_per_second": 74.833,
+      "eval_steps_per_second": 37.417,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.000925,
+      "loss": 0.9522,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.8565580248832703,
+      "eval_runtime": 4.0278,
+      "eval_samples_per_second": 74.483,
+      "eval_steps_per_second": 37.241,
+      "step": 1500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0009000000000000001,
+      "loss": 0.9143,
+      "step": 2000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.8810315132141113,
+      "eval_runtime": 4.0097,
+      "eval_samples_per_second": 74.819,
+      "eval_steps_per_second": 37.409,
+      "step": 2000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000875,
+      "loss": 0.9151,
+      "step": 2500
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.8461073040962219,
+      "eval_runtime": 4.0108,
+      "eval_samples_per_second": 74.798,
+      "eval_steps_per_second": 37.399,
+      "step": 2500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00085,
+      "loss": 0.9263,
+      "step": 3000
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.8521466255187988,
+      "eval_runtime": 4.0059,
+      "eval_samples_per_second": 74.89,
+      "eval_steps_per_second": 37.445,
+      "step": 3000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000825,
+      "loss": 0.9102,
+      "step": 3500
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.8521981835365295,
+      "eval_runtime": 4.0413,
+      "eval_samples_per_second": 74.234,
+      "eval_steps_per_second": 37.117,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0008,
+      "loss": 0.8866,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.8259674906730652,
+      "eval_runtime": 4.0316,
+      "eval_samples_per_second": 74.412,
+      "eval_steps_per_second": 37.206,
+      "step": 4000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0007750000000000001,
+      "loss": 0.911,
+      "step": 4500
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.8008900284767151,
+      "eval_runtime": 4.0224,
+      "eval_samples_per_second": 74.583,
+      "eval_steps_per_second": 37.292,
+      "step": 4500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00075,
+      "loss": 0.8536,
+      "step": 5000
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.808663547039032,
+      "eval_runtime": 5.0615,
+      "eval_samples_per_second": 59.271,
+      "eval_steps_per_second": 29.635,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 20000,
+  "num_train_epochs": 3,
+  "save_steps": 5000,
+  "total_flos": 8851794099093504.0,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-5000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0a404763c15e6a97a510a639844e059871b1160163baa499ef23d347aae03a98
-size 4091
+oid sha256:7a0a0ea9dbd2491eec13fb387c65610f8eb3edd4ca287eecb3bede1dd0d34916
+size 4155

config.json

@@ -3,6 +3,7 @@
   "architectures": [
     "T5ForConditionalGeneration"
   ],
+  "classifier_dropout": 0.0,
   "d_ff": 2048,
   "d_kv": 64,
   "d_model": 768,
@@ -55,7 +56,7 @@
   },
   "tie_word_embeddings": false,
   "torch_dtype": "float32",
-  "transformers_version": "4.31.0",
+  "transformers_version": "4.33.1",
   "use_cache": true,
   "vocab_size": 32128
 }

generation_config.json

@@ -3,5 +3,5 @@
   "decoder_start_token_id": 0,
   "eos_token_id": 1,
   "pad_token_id": 0,
-  "transformers_version": "4.31.0"
+  "transformers_version": "4.33.1"
 }

old/config.json

@@ -0,0 +1,61 @@
+{
+  "_name_or_path": "google/flan-t5-base",
+  "architectures": [
+    "T5ForConditionalGeneration"
+  ],
+  "d_ff": 2048,
+  "d_kv": 64,
+  "d_model": 768,
+  "decoder_start_token_id": 0,
+  "dense_act_fn": "gelu_new",
+  "dropout_rate": 0.1,
+  "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,
+  "model_type": "t5",
+  "n_positions": 512,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "relative_attention_max_distance": 128,
+  "relative_attention_num_buckets": 32,
+  "task_specific_params": {
+    "summarization": {
+      "early_stopping": true,
+      "length_penalty": 2.0,
+      "max_length": 200,
+      "min_length": 30,
+      "no_repeat_ngram_size": 3,
+      "num_beams": 4,
+      "prefix": "summarize: "
+    },
+    "translation_en_to_de": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to German: "
+    },
+    "translation_en_to_fr": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to French: "
+    },
+    "translation_en_to_ro": {
+      "early_stopping": true,
+      "max_length": 300,
+      "num_beams": 4,
+      "prefix": "translate English to Romanian: "
+    }
+  },
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.31.0",
+  "use_cache": true,
+  "vocab_size": 32128
+}

old/generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "decoder_start_token_id": 0,
+  "eos_token_id": 1,
+  "pad_token_id": 0,
+  "transformers_version": "4.31.0"
+}

old/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:852bf51b3b84310c6f60cef99cd64c1964d95b9ec68f08fd465090fbbb4fe39b
+size 2371333

old/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99c3909f22445cc0a707b7b83434ab5f95c2ab5bbddf1d7b9f4b14a3dddc8ff3
+size 990408885

old/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dba6c572f23aa9963da39e7213ee988824ff5da3fc545417fcd21bbf8c19dc45
+size 14575

old/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f5e24b8bf255cbefe9d307944a9741807d095b40cc5429a7befe9515b366b0f
+size 627

old/trainer_state.json

@@ -0,0 +1,296 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.5195263290501386,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.5e-05,
+      "loss": 0.8107,
+      "step": 500
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.6702454090118408,
+      "eval_runtime": 4.5337,
+      "eval_samples_per_second": 66.17,
+      "eval_steps_per_second": 33.085,
+      "step": 500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9e-05,
+      "loss": 0.717,
+      "step": 1000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.6662179827690125,
+      "eval_runtime": 5.109,
+      "eval_samples_per_second": 58.719,
+      "eval_steps_per_second": 29.36,
+      "step": 1000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 8.5e-05,
+      "loss": 0.7212,
+      "step": 1500
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.6485440135002136,
+      "eval_runtime": 4.6475,
+      "eval_samples_per_second": 64.551,
+      "eval_steps_per_second": 32.276,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 8e-05,
+      "loss": 0.7094,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.6365593671798706,
+      "eval_runtime": 4.9661,
+      "eval_samples_per_second": 60.41,
+      "eval_steps_per_second": 30.205,
+      "step": 2000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.7081,
+      "step": 2500
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.6411292552947998,
+      "eval_runtime": 5.0449,
+      "eval_samples_per_second": 59.466,
+      "eval_steps_per_second": 29.733,
+      "step": 2500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 7e-05,
+      "loss": 0.6803,
+      "step": 3000
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.629453718662262,
+      "eval_runtime": 5.1787,
+      "eval_samples_per_second": 57.929,
+      "eval_steps_per_second": 28.965,
+      "step": 3000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 0.6886,
+      "step": 3500
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.6171028017997742,
+      "eval_runtime": 6.0926,
+      "eval_samples_per_second": 49.24,
+      "eval_steps_per_second": 24.62,
+      "step": 3500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 6e-05,
+      "loss": 0.673,
+      "step": 4000
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 0.6199834942817688,
+      "eval_runtime": 4.5562,
+      "eval_samples_per_second": 65.844,
+      "eval_steps_per_second": 32.922,
+      "step": 4000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 0.6123,
+      "step": 4500
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 0.615871250629425,
+      "eval_runtime": 4.3914,
+      "eval_samples_per_second": 68.315,
+      "eval_steps_per_second": 34.158,
+      "step": 4500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.6036,
+      "step": 5000
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 0.6081268191337585,
+      "eval_runtime": 4.202,
+      "eval_samples_per_second": 71.395,
+      "eval_steps_per_second": 35.697,
+      "step": 5000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.5e-05,
+      "loss": 0.5854,
+      "step": 5500
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 0.610887885093689,
+      "eval_runtime": 4.5399,
+      "eval_samples_per_second": 66.08,
+      "eval_steps_per_second": 33.04,
+      "step": 5500
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4e-05,
+      "loss": 0.5936,
+      "step": 6000
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 0.6057543158531189,
+      "eval_runtime": 5.4963,
+      "eval_samples_per_second": 54.582,
+      "eval_steps_per_second": 27.291,
+      "step": 6000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.5e-05,
+      "loss": 0.5917,
+      "step": 6500
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 0.6038711071014404,
+      "eval_runtime": 4.4545,
+      "eval_samples_per_second": 67.348,
+      "eval_steps_per_second": 33.674,
+      "step": 6500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3e-05,
+      "loss": 0.5848,
+      "step": 7000
+    },
+    {
+      "epoch": 1.76,
+      "eval_loss": 0.6017142534255981,
+      "eval_runtime": 4.4106,
+      "eval_samples_per_second": 68.018,
+      "eval_steps_per_second": 34.009,
+      "step": 7000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.5e-05,
+      "loss": 0.5875,
+      "step": 7500
+    },
+    {
+      "epoch": 1.89,
+      "eval_loss": 0.6017369627952576,
+      "eval_runtime": 4.4904,
+      "eval_samples_per_second": 66.809,
+      "eval_steps_per_second": 33.404,
+      "step": 7500
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2e-05,
+      "loss": 0.5975,
+      "step": 8000
+    },
+    {
+      "epoch": 2.02,
+      "eval_loss": 0.5987575054168701,
+      "eval_runtime": 4.1962,
+      "eval_samples_per_second": 71.493,
+      "eval_steps_per_second": 35.747,
+      "step": 8000
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.5e-05,
+      "loss": 0.5551,
+      "step": 8500
+    },
+    {
+      "epoch": 2.14,
+      "eval_loss": 0.6018974781036377,
+      "eval_runtime": 4.1728,
+      "eval_samples_per_second": 71.894,
+      "eval_steps_per_second": 35.947,
+      "step": 8500
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1e-05,
+      "loss": 0.5338,
+      "step": 9000
+    },
+    {
+      "epoch": 2.27,
+      "eval_loss": 0.600771963596344,
+      "eval_runtime": 4.5111,
+      "eval_samples_per_second": 66.502,
+      "eval_steps_per_second": 33.251,
+      "step": 9000
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-06,
+      "loss": 0.5648,
+      "step": 9500
+    },
+    {
+      "epoch": 2.39,
+      "eval_loss": 0.599320650100708,
+      "eval_runtime": 4.3392,
+      "eval_samples_per_second": 69.138,
+      "eval_steps_per_second": 34.569,
+      "step": 9500
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.0,
+      "loss": 0.5328,
+      "step": 10000
+    },
+    {
+      "epoch": 2.52,
+      "eval_loss": 0.6002569794654846,
+      "eval_runtime": 5.4729,
+      "eval_samples_per_second": 54.816,
+      "eval_steps_per_second": 27.408,
+      "step": 10000
+    }
+  ],
+  "max_steps": 10000,
+  "num_train_epochs": 3,
+  "total_flos": 3.963545541632102e+16,
+  "trial_name": null,
+  "trial_params": null
+}

old/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a404763c15e6a97a510a639844e059871b1160163baa499ef23d347aae03a98
+size 4091

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:852bf51b3b84310c6f60cef99cd64c1964d95b9ec68f08fd465090fbbb4fe39b
+oid sha256:faf89a989d29c28e1adcc3475b1c824fe5329491ea1430eecc8d0e670a8fbcd3
 size 2371333

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:99c3909f22445cc0a707b7b83434ab5f95c2ab5bbddf1d7b9f4b14a3dddc8ff3
+oid sha256:8a68794e67b4a15b3adf98c8974799b4e08a27b56f9154e118dd392087fffa56
 size 990408885

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dba6c572f23aa9963da39e7213ee988824ff5da3fc545417fcd21bbf8c19dc45
+oid sha256:f1ff062526219ffde93f9f292e9365f17879a9afff6660d569e10c57e0a90df4
 size 14575

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3f5e24b8bf255cbefe9d307944a9741807d095b40cc5429a7befe9515b366b0f
+oid sha256:f58a0653df4cbbcf9d6cc03d846193b654e5e1cc8a7d6462c99377d7fbe445ea
 size 627

trainer_state.json

@@ -2,295 +2,578 @@
   "best_metric": null,
   "best_model_checkpoint": null,
   "epoch": 2.5195263290501386,
-  "global_step": 10000,
+  "eval_steps": 500,
+  "global_step": 20000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.13,
-      "learning_rate": 9.5e-05,
-      "loss": 0.8107,
+      "epoch": 0.06,
+      "learning_rate": 0.000975,
+      "loss": 1.0238,
       "step": 500
     },
     {
-      "epoch": 0.13,
-      "eval_loss": 0.6702454090118408,
-      "eval_runtime": 4.5337,
-      "eval_samples_per_second": 66.17,
-      "eval_steps_per_second": 33.085,
+      "epoch": 0.06,
+      "eval_loss": 0.8929525017738342,
+      "eval_runtime": 4.0089,
+      "eval_samples_per_second": 74.833,
+      "eval_steps_per_second": 37.417,
       "step": 500
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 9e-05,
-      "loss": 0.717,
+      "epoch": 0.13,
+      "learning_rate": 0.00095,
+      "loss": 0.9971,
       "step": 1000
     },
     {
-      "epoch": 0.25,
-      "eval_loss": 0.6662179827690125,
-      "eval_runtime": 5.109,
-      "eval_samples_per_second": 58.719,
-      "eval_steps_per_second": 29.36,
+      "epoch": 0.13,
+      "eval_loss": 0.8510345220565796,
+      "eval_runtime": 4.0089,
+      "eval_samples_per_second": 74.833,
+      "eval_steps_per_second": 37.417,
       "step": 1000
     },
     {
-      "epoch": 0.38,
-      "learning_rate": 8.5e-05,
-      "loss": 0.7212,
+      "epoch": 0.19,
+      "learning_rate": 0.000925,
+      "loss": 0.9522,
       "step": 1500
     },
     {
-      "epoch": 0.38,
-      "eval_loss": 0.6485440135002136,
-      "eval_runtime": 4.6475,
-      "eval_samples_per_second": 64.551,
-      "eval_steps_per_second": 32.276,
+      "epoch": 0.19,
+      "eval_loss": 0.8565580248832703,
+      "eval_runtime": 4.0278,
+      "eval_samples_per_second": 74.483,
+      "eval_steps_per_second": 37.241,
       "step": 1500
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 8e-05,
-      "loss": 0.7094,
+      "epoch": 0.25,
+      "learning_rate": 0.0009000000000000001,
+      "loss": 0.9143,
       "step": 2000
     },
     {
-      "epoch": 0.5,
-      "eval_loss": 0.6365593671798706,
-      "eval_runtime": 4.9661,
-      "eval_samples_per_second": 60.41,
-      "eval_steps_per_second": 30.205,
+      "epoch": 0.25,
+      "eval_loss": 0.8810315132141113,
+      "eval_runtime": 4.0097,
+      "eval_samples_per_second": 74.819,
+      "eval_steps_per_second": 37.409,
       "step": 2000
     },
     {
-      "epoch": 0.63,
-      "learning_rate": 7.500000000000001e-05,
-      "loss": 0.7081,
+      "epoch": 0.31,
+      "learning_rate": 0.000875,
+      "loss": 0.9151,
       "step": 2500
     },
     {
-      "epoch": 0.63,
-      "eval_loss": 0.6411292552947998,
-      "eval_runtime": 5.0449,
-      "eval_samples_per_second": 59.466,
-      "eval_steps_per_second": 29.733,
+      "epoch": 0.31,
+      "eval_loss": 0.8461073040962219,
+      "eval_runtime": 4.0108,
+      "eval_samples_per_second": 74.798,
+      "eval_steps_per_second": 37.399,
       "step": 2500
     },
     {
-      "epoch": 0.76,
-      "learning_rate": 7e-05,
-      "loss": 0.6803,
+      "epoch": 0.38,
+      "learning_rate": 0.00085,
+      "loss": 0.9263,
       "step": 3000
     },
     {
-      "epoch": 0.76,
-      "eval_loss": 0.629453718662262,
-      "eval_runtime": 5.1787,
-      "eval_samples_per_second": 57.929,
-      "eval_steps_per_second": 28.965,
+      "epoch": 0.38,
+      "eval_loss": 0.8521466255187988,
+      "eval_runtime": 4.0059,
+      "eval_samples_per_second": 74.89,
+      "eval_steps_per_second": 37.445,
       "step": 3000
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 6.500000000000001e-05,
-      "loss": 0.6886,
+      "epoch": 0.44,
+      "learning_rate": 0.000825,
+      "loss": 0.9102,
       "step": 3500
     },
     {
-      "epoch": 0.88,
-      "eval_loss": 0.6171028017997742,
-      "eval_runtime": 6.0926,
-      "eval_samples_per_second": 49.24,
-      "eval_steps_per_second": 24.62,
+      "epoch": 0.44,
+      "eval_loss": 0.8521981835365295,
+      "eval_runtime": 4.0413,
+      "eval_samples_per_second": 74.234,
+      "eval_steps_per_second": 37.117,
       "step": 3500
     },
     {
-      "epoch": 1.01,
-      "learning_rate": 6e-05,
-      "loss": 0.673,
+      "epoch": 0.5,
+      "learning_rate": 0.0008,
+      "loss": 0.8866,
       "step": 4000
     },
     {
-      "epoch": 1.01,
-      "eval_loss": 0.6199834942817688,
-      "eval_runtime": 4.5562,
-      "eval_samples_per_second": 65.844,
-      "eval_steps_per_second": 32.922,
+      "epoch": 0.5,
+      "eval_loss": 0.8259674906730652,
+      "eval_runtime": 4.0316,
+      "eval_samples_per_second": 74.412,
+      "eval_steps_per_second": 37.206,
       "step": 4000
     },
     {
-      "epoch": 1.13,
-      "learning_rate": 5.500000000000001e-05,
-      "loss": 0.6123,
+      "epoch": 0.57,
+      "learning_rate": 0.0007750000000000001,
+      "loss": 0.911,
       "step": 4500
     },
     {
-      "epoch": 1.13,
-      "eval_loss": 0.615871250629425,
-      "eval_runtime": 4.3914,
-      "eval_samples_per_second": 68.315,
-      "eval_steps_per_second": 34.158,
+      "epoch": 0.57,
+      "eval_loss": 0.8008900284767151,
+      "eval_runtime": 4.0224,
+      "eval_samples_per_second": 74.583,
+      "eval_steps_per_second": 37.292,
       "step": 4500
     },
     {
-      "epoch": 1.26,
-      "learning_rate": 5e-05,
-      "loss": 0.6036,
+      "epoch": 0.63,
+      "learning_rate": 0.00075,
+      "loss": 0.8536,
       "step": 5000
     },
     {
-      "epoch": 1.26,
-      "eval_loss": 0.6081268191337585,
-      "eval_runtime": 4.202,
-      "eval_samples_per_second": 71.395,
-      "eval_steps_per_second": 35.697,
+      "epoch": 0.63,
+      "eval_loss": 0.808663547039032,
+      "eval_runtime": 5.0615,
+      "eval_samples_per_second": 59.271,
+      "eval_steps_per_second": 29.635,
       "step": 5000
     },
     {
-      "epoch": 1.39,
-      "learning_rate": 4.5e-05,
-      "loss": 0.5854,
+      "epoch": 0.69,
+      "learning_rate": 0.000725,
+      "loss": 0.8277,
       "step": 5500
     },
     {
-      "epoch": 1.39,
-      "eval_loss": 0.610887885093689,
-      "eval_runtime": 4.5399,
-      "eval_samples_per_second": 66.08,
-      "eval_steps_per_second": 33.04,
+      "epoch": 0.69,
+      "eval_loss": 0.7903586626052856,
+      "eval_runtime": 4.0081,
+      "eval_samples_per_second": 74.849,
+      "eval_steps_per_second": 37.425,
       "step": 5500
     },
     {
-      "epoch": 1.51,
-      "learning_rate": 4e-05,
-      "loss": 0.5936,
+      "epoch": 0.76,
+      "learning_rate": 0.0007,
+      "loss": 0.8413,
       "step": 6000
     },
     {
-      "epoch": 1.51,
-      "eval_loss": 0.6057543158531189,
-      "eval_runtime": 5.4963,
-      "eval_samples_per_second": 54.582,
-      "eval_steps_per_second": 27.291,
+      "epoch": 0.76,
+      "eval_loss": 0.8013813495635986,
+      "eval_runtime": 3.994,
+      "eval_samples_per_second": 75.113,
+      "eval_steps_per_second": 37.556,
       "step": 6000
     },
     {
-      "epoch": 1.64,
-      "learning_rate": 3.5e-05,
-      "loss": 0.5917,
+      "epoch": 0.82,
+      "learning_rate": 0.000675,
+      "loss": 0.8491,
       "step": 6500
     },
     {
-      "epoch": 1.64,
-      "eval_loss": 0.6038711071014404,
-      "eval_runtime": 4.4545,
-      "eval_samples_per_second": 67.348,
-      "eval_steps_per_second": 33.674,
+      "epoch": 0.82,
+      "eval_loss": 0.7867017984390259,
+      "eval_runtime": 4.0072,
+      "eval_samples_per_second": 74.864,
+      "eval_steps_per_second": 37.432,
       "step": 6500
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 3e-05,
-      "loss": 0.5848,
+      "epoch": 0.88,
+      "learning_rate": 0.0006500000000000001,
+      "loss": 0.8077,
       "step": 7000
     },
     {
-      "epoch": 1.76,
-      "eval_loss": 0.6017142534255981,
-      "eval_runtime": 4.4106,
-      "eval_samples_per_second": 68.018,
-      "eval_steps_per_second": 34.009,
+      "epoch": 0.88,
+      "eval_loss": 0.759165346622467,
+      "eval_runtime": 4.0166,
+      "eval_samples_per_second": 74.691,
+      "eval_steps_per_second": 37.345,
       "step": 7000
     },
     {
-      "epoch": 1.89,
-      "learning_rate": 2.5e-05,
-      "loss": 0.5875,
+      "epoch": 0.94,
+      "learning_rate": 0.000625,
+      "loss": 0.796,
       "step": 7500
     },
     {
-      "epoch": 1.89,
-      "eval_loss": 0.6017369627952576,
-      "eval_runtime": 4.4904,
-      "eval_samples_per_second": 66.809,
-      "eval_steps_per_second": 33.404,
+      "epoch": 0.94,
+      "eval_loss": 0.7527943849563599,
+      "eval_runtime": 3.998,
+      "eval_samples_per_second": 75.037,
+      "eval_steps_per_second": 37.518,
       "step": 7500
     },
     {
-      "epoch": 2.02,
-      "learning_rate": 2e-05,
-      "loss": 0.5975,
+      "epoch": 1.01,
+      "learning_rate": 0.0006,
+      "loss": 0.7952,
       "step": 8000
     },
     {
-      "epoch": 2.02,
-      "eval_loss": 0.5987575054168701,
-      "eval_runtime": 4.1962,
-      "eval_samples_per_second": 71.493,
-      "eval_steps_per_second": 35.747,
+      "epoch": 1.01,
+      "eval_loss": 0.7407393455505371,
+      "eval_runtime": 3.9421,
+      "eval_samples_per_second": 76.102,
+      "eval_steps_per_second": 38.051,
       "step": 8000
     },
     {
-      "epoch": 2.14,
-      "learning_rate": 1.5e-05,
-      "loss": 0.5551,
+      "epoch": 1.07,
+      "learning_rate": 0.000575,
+      "loss": 0.5761,
       "step": 8500
     },
     {
-      "epoch": 2.14,
-      "eval_loss": 0.6018974781036377,
-      "eval_runtime": 4.1728,
-      "eval_samples_per_second": 71.894,
-      "eval_steps_per_second": 35.947,
+      "epoch": 1.07,
+      "eval_loss": 0.7574229836463928,
+      "eval_runtime": 3.9855,
+      "eval_samples_per_second": 75.273,
+      "eval_steps_per_second": 37.636,
       "step": 8500
     },
     {
-      "epoch": 2.27,
-      "learning_rate": 1e-05,
-      "loss": 0.5338,
+      "epoch": 1.13,
+      "learning_rate": 0.00055,
+      "loss": 0.6084,
       "step": 9000
     },
     {
-      "epoch": 2.27,
-      "eval_loss": 0.600771963596344,
-      "eval_runtime": 4.5111,
-      "eval_samples_per_second": 66.502,
-      "eval_steps_per_second": 33.251,
+      "epoch": 1.13,
+      "eval_loss": 0.7349967956542969,
+      "eval_runtime": 3.995,
+      "eval_samples_per_second": 75.093,
+      "eval_steps_per_second": 37.547,
       "step": 9000
     },
     {
-      "epoch": 2.39,
-      "learning_rate": 5e-06,
-      "loss": 0.5648,
+      "epoch": 1.2,
+      "learning_rate": 0.0005250000000000001,
+      "loss": 0.5863,
       "step": 9500
     },
     {
-      "epoch": 2.39,
-      "eval_loss": 0.599320650100708,
-      "eval_runtime": 4.3392,
-      "eval_samples_per_second": 69.138,
-      "eval_steps_per_second": 34.569,
+      "epoch": 1.2,
+      "eval_loss": 0.7407773733139038,
+      "eval_runtime": 3.9874,
+      "eval_samples_per_second": 75.238,
+      "eval_steps_per_second": 37.619,
       "step": 9500
     },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0005,
+      "loss": 0.5835,
+      "step": 10000
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 0.719527006149292,
+      "eval_runtime": 3.9471,
+      "eval_samples_per_second": 76.006,
+      "eval_steps_per_second": 38.003,
+      "step": 10000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.000475,
+      "loss": 0.5751,
+      "step": 10500
+    },
+    {
+      "epoch": 1.32,
+      "eval_loss": 0.7423205971717834,
+      "eval_runtime": 3.9978,
+      "eval_samples_per_second": 75.041,
+      "eval_steps_per_second": 37.52,
+      "step": 10500
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00045000000000000004,
+      "loss": 0.5746,
+      "step": 11000
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 0.7284040451049805,
+      "eval_runtime": 3.9657,
+      "eval_samples_per_second": 75.649,
+      "eval_steps_per_second": 37.824,
+      "step": 11000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.000425,
+      "loss": 0.5847,
+      "step": 11500
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 0.7247716188430786,
+      "eval_runtime": 3.962,
+      "eval_samples_per_second": 75.719,
+      "eval_steps_per_second": 37.86,
+      "step": 11500
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0004,
+      "loss": 0.5759,
+      "step": 12000
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 0.7151039838790894,
+      "eval_runtime": 3.9922,
+      "eval_samples_per_second": 75.147,
+      "eval_steps_per_second": 37.573,
+      "step": 12000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.000375,
+      "loss": 0.5922,
+      "step": 12500
+    },
+    {
+      "epoch": 1.57,
+      "eval_loss": 0.7031030058860779,
+      "eval_runtime": 4.0386,
+      "eval_samples_per_second": 74.282,
+      "eval_steps_per_second": 37.141,
+      "step": 12500
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00035,
+      "loss": 0.5678,
+      "step": 13000
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 0.710035502910614,
+      "eval_runtime": 4.0808,
+      "eval_samples_per_second": 73.515,
+      "eval_steps_per_second": 36.758,
+      "step": 13000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00032500000000000004,
+      "loss": 0.5627,
+      "step": 13500
+    },
+    {
+      "epoch": 1.7,
+      "eval_loss": 0.7093074321746826,
+      "eval_runtime": 4.0326,
+      "eval_samples_per_second": 74.393,
+      "eval_steps_per_second": 37.197,
+      "step": 13500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0003,
+      "loss": 0.5568,
+      "step": 14000
+    },
+    {
+      "epoch": 1.76,
+      "eval_loss": 0.6942310333251953,
+      "eval_runtime": 4.0875,
+      "eval_samples_per_second": 73.395,
+      "eval_steps_per_second": 36.697,
+      "step": 14000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.000275,
+      "loss": 0.5589,
+      "step": 14500
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 0.6947352290153503,
+      "eval_runtime": 3.9856,
+      "eval_samples_per_second": 75.27,
+      "eval_steps_per_second": 37.635,
+      "step": 14500
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00025,
+      "loss": 0.5675,
+      "step": 15000
+    },
+    {
+      "epoch": 1.89,
+      "eval_loss": 0.6902616620063782,
+      "eval_runtime": 4.0066,
+      "eval_samples_per_second": 74.876,
+      "eval_steps_per_second": 37.438,
+      "step": 15000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00022500000000000002,
+      "loss": 0.5422,
+      "step": 15500
+    },
+    {
+      "epoch": 1.95,
+      "eval_loss": 0.6929380297660828,
+      "eval_runtime": 4.107,
+      "eval_samples_per_second": 73.047,
+      "eval_steps_per_second": 36.523,
+      "step": 15500
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.0002,
+      "loss": 0.5104,
+      "step": 16000
+    },
+    {
+      "epoch": 2.02,
+      "eval_loss": 0.7098422050476074,
+      "eval_runtime": 4.0323,
+      "eval_samples_per_second": 74.4,
+      "eval_steps_per_second": 37.2,
+      "step": 16000
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.000175,
+      "loss": 0.3835,
+      "step": 16500
+    },
+    {
+      "epoch": 2.08,
+      "eval_loss": 0.7105218768119812,
+      "eval_runtime": 4.0594,
+      "eval_samples_per_second": 73.903,
+      "eval_steps_per_second": 36.952,
+      "step": 16500
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.00015,
+      "loss": 0.3805,
+      "step": 17000
+    },
+    {
+      "epoch": 2.14,
+      "eval_loss": 0.7144222855567932,
+      "eval_runtime": 4.0853,
+      "eval_samples_per_second": 73.434,
+      "eval_steps_per_second": 36.717,
+      "step": 17000
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.000125,
+      "loss": 0.3718,
+      "step": 17500
+    },
+    {
+      "epoch": 2.2,
+      "eval_loss": 0.7210414409637451,
+      "eval_runtime": 5.0511,
+      "eval_samples_per_second": 59.393,
+      "eval_steps_per_second": 29.697,
+      "step": 17500
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.0001,
+      "loss": 0.3688,
+      "step": 18000
+    },
+    {
+      "epoch": 2.27,
+      "eval_loss": 0.7145898342132568,
+      "eval_runtime": 4.7793,
+      "eval_samples_per_second": 62.77,
+      "eval_steps_per_second": 31.385,
+      "step": 18000
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 7.5e-05,
+      "loss": 0.3645,
+      "step": 18500
+    },
+    {
+      "epoch": 2.33,
+      "eval_loss": 0.7136221528053284,
+      "eval_runtime": 4.0171,
+      "eval_samples_per_second": 74.681,
+      "eval_steps_per_second": 37.34,
+      "step": 18500
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.3865,
+      "step": 19000
+    },
+    {
+      "epoch": 2.39,
+      "eval_loss": 0.7075753808021545,
+      "eval_runtime": 3.9658,
+      "eval_samples_per_second": 75.646,
+      "eval_steps_per_second": 37.823,
+      "step": 19000
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.5e-05,
+      "loss": 0.3633,
+      "step": 19500
+    },
+    {
+      "epoch": 2.46,
+      "eval_loss": 0.7097809314727783,
+      "eval_runtime": 4.0163,
+      "eval_samples_per_second": 74.696,
+      "eval_steps_per_second": 37.348,
+      "step": 19500
+    },
     {
       "epoch": 2.52,
       "learning_rate": 0.0,
-      "loss": 0.5328,
-      "step": 10000
+      "loss": 0.3674,
+      "step": 20000
     },
     {
       "epoch": 2.52,
-      "eval_loss": 0.6002569794654846,
-      "eval_runtime": 5.4729,
-      "eval_samples_per_second": 54.816,
-      "eval_steps_per_second": 27.408,
-      "step": 10000
+      "eval_loss": 0.7079904079437256,
+      "eval_runtime": 4.0804,
+      "eval_samples_per_second": 73.522,
+      "eval_steps_per_second": 36.761,
+      "step": 20000
     }
   ],
-  "max_steps": 10000,
+  "logging_steps": 500,
+  "max_steps": 20000,
   "num_train_epochs": 3,
-  "total_flos": 3.963545541632102e+16,
+  "save_steps": 5000,
+  "total_flos": 3.549121832463667e+16,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0a404763c15e6a97a510a639844e059871b1160163baa499ef23d347aae03a98
-size 4091
+oid sha256:7a0a0ea9dbd2491eec13fb387c65610f8eb3edd4ca287eecb3bede1dd0d34916
+size 4155