Restore

.gitignore

@@ -0,0 +1 @@
+checkpoint-*/

README.md

@@ -0,0 +1,77 @@
+---
+language:
+- en
+license: apache-2.0
+tags:
+- generated_from_trainer
+datasets:
+- top_v2
+model-index:
+- name: t5-base-pointer-adv-top_v2
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# t5-base-pointer-adv-top_v2
+
+This model is a fine-tuned version of [google/mt5-base](https://huggingface.co/google/mt5-base) on the top_v2 dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0255
+- Exact Match: 0.8472
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 0.001
+- train_batch_size: 8
+- eval_batch_size: 8
+- seed: 42
+- gradient_accumulation_steps: 64
+- total_train_batch_size: 512
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- training_steps: 3000
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | Exact Match |
+|:-------------:|:-----:|:----:|:---------------:|:-----------:|
+| 2.2938        | 0.41  | 200  | 0.5532          | 0.0012      |
+| 0.671         | 0.82  | 400  | 0.1624          | 0.1610      |
+| 0.5276        | 1.23  | 600  | 0.0692          | 0.2157      |
+| 0.4196        | 1.64  | 800  | 0.0491          | 0.2259      |
+| 0.3593        | 2.05  | 1000 | 0.0400          | 0.2291      |
+| 0.3471        | 2.46  | 1200 | 0.0335          | 0.2297      |
+| 0.3416        | 2.87  | 1400 | 0.0307          | 0.2318      |
+| 0.3351        | 3.29  | 1600 | 0.0307          | 0.2334      |
+| 0.3316        | 3.7   | 1800 | 0.0297          | 0.2343      |
+| 0.3312        | 4.11  | 2000 | 0.0282          | 0.2344      |
+| 0.3271        | 4.52  | 2200 | 0.0262          | 0.2365      |
+| 0.3241        | 4.93  | 2400 | 0.0263          | 0.2365      |
+| 0.3227        | 5.34  | 2600 | 0.0259          | 0.2368      |
+| 0.3201        | 5.75  | 2800 | 0.0257          | 0.2365      |
+| 0.3227        | 6.16  | 3000 | 0.0255          | 0.2365      |
+
+
+### Framework versions
+
+- Transformers 4.24.0
+- Pytorch 1.13.0+cu117
+- Datasets 2.7.0
+- Tokenizers 0.13.2

added_tokens.json

@@ -0,0 +1,224 @@
+{
+  "<pt-0>": 250104,
+  "<pt-10>": 250111,
+  "<pt-11>": 250134,
+  "<pt-12>": 250128,
+  "<pt-13>": 250132,
+  "<pt-14>": 250103,
+  "<pt-15>": 250120,
+  "<pt-16>": 250117,
+  "<pt-17>": 250131,
+  "<pt-18>": 250116,
+  "<pt-19>": 250106,
+  "<pt-1>": 250122,
+  "<pt-20>": 250119,
+  "<pt-21>": 250136,
+  "<pt-22>": 250137,
+  "<pt-23>": 250139,
+  "<pt-24>": 250141,
+  "<pt-25>": 250140,
+  "<pt-26>": 250170,
+  "<pt-27>": 250177,
+  "<pt-28>": 250179,
+  "<pt-29>": 250193,
+  "<pt-2>": 250105,
+  "<pt-30>": 250183,
+  "<pt-31>": 250188,
+  "<pt-32>": 250167,
+  "<pt-33>": 250194,
+  "<pt-34>": 250180,
+  "<pt-35>": 250187,
+  "<pt-36>": 250166,
+  "<pt-37>": 250176,
+  "<pt-38>": 250186,
+  "<pt-39>": 250163,
+  "<pt-3>": 250124,
+  "<pt-40>": 250175,
+  "<pt-41>": 250184,
+  "<pt-42>": 250164,
+  "<pt-43>": 250181,
+  "<pt-44>": 250182,
+  "<pt-45>": 250189,
+  "<pt-46>": 250165,
+  "<pt-47>": 250190,
+  "<pt-48>": 250192,
+  "<pt-49>": 250173,
+  "<pt-4>": 250121,
+  "<pt-50>": 250185,
+  "<pt-51>": 250178,
+  "<pt-52>": 250191,
+  "<pt-53>": 250168,
+  "<pt-54>": 250174,
+  "<pt-55>": 250169,
+  "<pt-5>": 250115,
+  "<pt-6>": 250112,
+  "<pt-7>": 250113,
+  "<pt-8>": 250123,
+  "<pt-9>": 250108,
+  "[IN:ADD_TIME_TIMER": 250308,
+  "[IN:ADD_TO_PLAYLIST_MUSIC": 250222,
+  "[IN:CANCEL_MESSAGE": 250201,
+  "[IN:CREATE_ALARM": 250130,
+  "[IN:CREATE_PLAYLIST_MUSIC": 250223,
+  "[IN:CREATE_REMINDER": 250279,
+  "[IN:CREATE_TIMER": 250303,
+  "[IN:DELETE_ALARM": 250127,
+  "[IN:DELETE_REMINDER": 250274,
+  "[IN:DELETE_TIMER": 250299,
+  "[IN:DISLIKE_MUSIC": 250226,
+  "[IN:GET_ALARM": 250109,
+  "[IN:GET_BIRTHDAY": 250296,
+  "[IN:GET_CONTACT": 250149,
+  "[IN:GET_DIRECTIONS": 250266,
+  "[IN:GET_DISTANCE": 250245,
+  "[IN:GET_ESTIMATED_ARRIVAL": 250255,
+  "[IN:GET_ESTIMATED_DEPARTURE": 250262,
+  "[IN:GET_ESTIMATED_DURATION": 250267,
+  "[IN:GET_EVENT": 250160,
+  "[IN:GET_EVENT_ATTENDEE": 250162,
+  "[IN:GET_EVENT_ATTENDEE_AMOUNT": 250195,
+  "[IN:GET_EVENT_ORGANIZER": 250197,
+  "[IN:GET_INFO_CONTACT": 250317,
+  "[IN:GET_INFO_ROAD_CONDITION": 250258,
+  "[IN:GET_INFO_ROUTE": 250244,
+  "[IN:GET_INFO_TRAFFIC": 250257,
+  "[IN:GET_LOCATION": 250143,
+  "[IN:GET_LOCATION_HOME": 250150,
+  "[IN:GET_LOCATION_HOMETOWN": 250268,
+  "[IN:GET_LOCATION_SCHOOL": 250148,
+  "[IN:GET_LOCATION_WORK": 250171,
+  "[IN:GET_MESSAGE": 250209,
+  "[IN:GET_RECURRING_DATE_TIME": 250270,
+  "[IN:GET_REMINDER": 250273,
+  "[IN:GET_REMINDER_AMOUNT": 250291,
+  "[IN:GET_REMINDER_DATE_TIME": 250288,
+  "[IN:GET_REMINDER_LOCATION": 250290,
+  "[IN:GET_SUNRISE": 250314,
+  "[IN:GET_SUNSET": 250310,
+  "[IN:GET_TIME": 250101,
+  "[IN:GET_TIMER": 250306,
+  "[IN:GET_TODO": 250278,
+  "[IN:GET_WEATHER": 250311,
+  "[IN:HELP_REMINDER": 250292,
+  "[IN:IGNORE_MESSAGE": 250218,
+  "[IN:LIKE_MUSIC": 250240,
+  "[IN:LOOP_MUSIC": 250237,
+  "[IN:NEGATION": 250196,
+  "[IN:PAUSE_MUSIC": 250221,
+  "[IN:PAUSE_TIMER": 250304,
+  "[IN:PLAY_MUSIC": 250231,
+  "[IN:PREVIOUS_TRACK_MUSIC": 250234,
+  "[IN:REACT_MESSAGE": 250210,
+  "[IN:REMOVE_FROM_PLAYLIST_MUSIC": 250236,
+  "[IN:REPLAY_MUSIC": 250235,
+  "[IN:REPLY_MESSAGE": 250283,
+  "[IN:RESTART_TIMER": 250302,
+  "[IN:RESUME_TIMER": 250309,
+  "[IN:SELECT_ITEM": 250320,
+  "[IN:SEND_MESSAGE": 250205,
+  "[IN:SEND_TEXT_MESSAGE": 250219,
+  "[IN:SET_DEFAULT_PROVIDER_MUSIC": 250321,
+  "[IN:SILENCE_ALARM": 250107,
+  "[IN:SKIP_TRACK_MUSIC": 250239,
+  "[IN:SNOOZE_ALARM": 250114,
+  "[IN:START_SHUFFLE_MUSIC": 250224,
+  "[IN:STOP_MUSIC": 250229,
+  "[IN:SUBTRACT_TIME_TIMER": 250298,
+  "[IN:UNSUPPORTED_ALARM": 250102,
+  "[IN:UNSUPPORTED_EVENT": 250159,
+  "[IN:UNSUPPORTED_MESSAGING": 250217,
+  "[IN:UNSUPPORTED_MUSIC": 250230,
+  "[IN:UNSUPPORTED_NAVIGATION": 250251,
+  "[IN:UNSUPPORTED_TIMER": 250301,
+  "[IN:UNSUPPORTED_WEATHER": 250313,
+  "[IN:UPDATE_ALARM": 250110,
+  "[IN:UPDATE_DIRECTIONS": 250248,
+  "[IN:UPDATE_REMINDER": 250285,
+  "[IN:UPDATE_REMINDER_DATE_TIME": 250275,
+  "[IN:UPDATE_REMINDER_TODO": 250277,
+  "[IN:UPDATE_TIMER": 250307,
+  "[SL:AGE": 250220,
+  "[SL:ALARM_NAME": 250118,
+  "[SL:AMOUNT": 250135,
+  "[SL:ATTENDEE": 250276,
+  "[SL:ATTENDEE_ADDED": 250295,
+  "[SL:ATTENDEE_EVENT": 250152,
+  "[SL:ATTENDEE_REMOVED": 250294,
+  "[SL:ATTRIBUTE_EVENT": 250157,
+  "[SL:BIRTHDAY": 250319,
+  "[SL:CATEGORY_EVENT": 250156,
+  "[SL:CATEGORY_LOCATION": 250151,
+  "[SL:CONTACT": 250154,
+  "[SL:CONTACT_RELATED": 250146,
+  "[SL:CONTENT_EMOJI": 250208,
+  "[SL:CONTENT_EXACT": 250198,
+  "[SL:DATE_TIME": 250126,
+  "[SL:DATE_TIME_ARRIVAL": 250261,
+  "[SL:DATE_TIME_BIRTHDAY": 250318,
+  "[SL:DATE_TIME_DEPARTURE": 250263,
+  "[SL:DATE_TIME_NEW": 250272,
+  "[SL:DATE_TIME_RECURRING": 250100,
+  "[SL:DESTINATION": 250246,
+  "[SL:DURATION": 250133,
+  "[SL:FREQUENCY": 250282,
+  "[SL:GROUP": 250203,
+  "[SL:JOB": 250297,
+  "[SL:LOCATION": 250161,
+  "[SL:LOCATION_CURRENT": 250264,
+  "[SL:LOCATION_HOME": 250206,
+  "[SL:LOCATION_MODIFIER": 250155,
+  "[SL:LOCATION_USER": 250144,
+  "[SL:LOCATION_WORK": 250260,
+  "[SL:MEASUREMENT_UNIT": 250316,
+  "[SL:METHOD_RETRIEVAL_REMINDER": 250281,
+  "[SL:METHOD_TIMER": 250305,
+  "[SL:METHOD_TRAVEL": 250253,
+  "[SL:MUSIC_ALBUM_TITLE": 250233,
+  "[SL:MUSIC_ARTIST_NAME": 250225,
+  "[SL:MUSIC_GENRE": 250228,
+  "[SL:MUSIC_PLAYLIST_TITLE": 250241,
+  "[SL:MUSIC_PROVIDER_NAME": 250232,
+  "[SL:MUSIC_RADIO_ID": 250242,
+  "[SL:MUSIC_TRACK_TITLE": 250238,
+  "[SL:MUSIC_TYPE": 250227,
+  "[SL:MUTUAL_EMPLOYER": 250215,
+  "[SL:MUTUAL_LOCATION": 250212,
+  "[SL:MUTUAL_SCHOOL": 250214,
+  "[SL:NAME_APP": 250293,
+  "[SL:NAME_EVENT": 250158,
+  "[SL:OBSTRUCTION_AVOID": 250254,
+  "[SL:ORDINAL": 250125,
+  "[SL:ORGANIZER_EVENT": 250172,
+  "[SL:PATH": 250243,
+  "[SL:PATH_AVOID": 250250,
+  "[SL:PERIOD": 250129,
+  "[SL:PERSON_REMINDED": 250284,
+  "[SL:PERSON_REMINDED_ADDED": 250287,
+  "[SL:PERSON_REMINDED_REMOVED": 250289,
+  "[SL:POINT_ON_MAP": 250145,
+  "[SL:RECIPIENT": 250207,
+  "[SL:RECURRING_DATE_TIME": 250142,
+  "[SL:RECURRING_DATE_TIME_NEW": 250286,
+  "[SL:RESOURCE": 250213,
+  "[SL:ROAD_CONDITION": 250249,
+  "[SL:ROAD_CONDITION_AVOID": 250265,
+  "[SL:SEARCH_RADIUS": 250153,
+  "[SL:SENDER": 250199,
+  "[SL:SOURCE": 250247,
+  "[SL:TAG_MESSAGE": 250216,
+  "[SL:TIMER_NAME": 250300,
+  "[SL:TIME_ZONE": 250138,
+  "[SL:TODO": 250271,
+  "[SL:TODO_NEW": 250280,
+  "[SL:TYPE_CONTACT": 250204,
+  "[SL:TYPE_CONTENT": 250200,
+  "[SL:TYPE_INFO": 250211,
+  "[SL:TYPE_REACTION": 250202,
+  "[SL:TYPE_RELATION": 250147,
+  "[SL:UNIT_DISTANCE": 250259,
+  "[SL:WAYPOINT": 250256,
+  "[SL:WAYPOINT_ADDED": 250269,
+  "[SL:WAYPOINT_AVOID": 250252,
+  "[SL:WEATHER_ATTRIBUTE": 250312,
+  "[SL:WEATHER_TEMPERATURE_UNIT": 250315
+}

all_results.json

@@ -0,0 +1,20 @@
+{
+    "epoch": 6.16,
+    "eval_exact_match": 0.8472027972027972,
+    "eval_loss": 0.02550993673503399,
+    "eval_runtime": 1857.16,
+    "eval_samples": 17160,
+    "eval_samples_per_second": 9.24,
+    "eval_steps_per_second": 1.155,
+    "predict_exact_match": 0.8496583730823772,
+    "predict_loss": 0.028981352224946022,
+    "predict_runtime": 4150.6435,
+    "predict_samples": 38785,
+    "predict_samples_per_second": 9.344,
+    "predict_steps_per_second": 1.168,
+    "train_loss": 0.5049722137451171,
+    "train_runtime": 47265.062,
+    "train_samples": 249508,
+    "train_samples_per_second": 32.498,
+    "train_steps_per_second": 0.063
+}

config.json

@@ -0,0 +1,32 @@
+{
+  "_name_or_path": "google/mt5-base",
+  "architectures": [
+    "AlignedMT5ForConditionalGeneration"
+  ],
+  "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": "mt5",
+  "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,
+  "tie_word_embeddings": false,
+  "tokenizer_class": "T5Tokenizer",
+  "torch_dtype": "float32",
+  "transformers_version": "4.24.0",
+  "use_cache": true,
+  "vocab_size": 250322
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 6.16,
+    "eval_exact_match": 0.8472027972027972,
+    "eval_loss": 0.02550993673503399,
+    "eval_runtime": 1857.16,
+    "eval_samples": 17160,
+    "eval_samples_per_second": 9.24,
+    "eval_steps_per_second": 1.155
+}

predict_results.json

@@ -0,0 +1,8 @@
+{
+    "predict_exact_match": 0.8496583730823772,
+    "predict_loss": 0.028981352224946022,
+    "predict_runtime": 4150.6435,
+    "predict_samples": 38785,
+    "predict_samples_per_second": 9.344,
+    "predict_steps_per_second": 1.168
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0a9bf5410c26b017552fcf3256b41ebc72e68748a195c57f9ab65dea5b86073
+size 2333360985

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:e1d9bf8581400dbadfa0f7c4f9c061fb88fa5afdffd47848f883a217b60a6920
+size 16373140

tokenizer_config.json

@@ -0,0 +1,11 @@
+{
+  "additional_special_tokens": null,
+  "eos_token": "</s>",
+  "extra_ids": 0,
+  "name_or_path": "google/mt5-base",
+  "pad_token": "<pad>",
+  "sp_model_kwargs": {},
+  "special_tokens_map_file": "/home/patrick/.cache/torch/transformers/685ac0ca8568ec593a48b61b0a3c272beee9bc194a3c7241d15dcadb5f875e53.f76030f3ec1b96a8199b2593390c610e76ca8028ef3d24680000619ffb646276",
+  "tokenizer_class": "T5Tokenizer",
+  "unk_token": "<unk>"
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 6.16,
+    "train_loss": 0.5049722137451171,
+    "train_runtime": 47265.062,
+    "train_samples": 249508,
+    "train_samples_per_second": 32.498,
+    "train_steps_per_second": 0.063
+}

trainer_state.json

@@ -0,0 +1,250 @@
+{
+  "best_metric": 0.02550993673503399,
+  "best_model_checkpoint": "/data/wheld3/mt5-base-pointer-adv-top_v2/checkpoint-3000",
+  "epoch": 6.160056430151656,
+  "global_step": 3000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0009333333333333333,
+      "loss": 2.2938,
+      "step": 200
+    },
+    {
+      "epoch": 0.41,
+      "eval_exact_match": 0.0011655011655011655,
+      "eval_loss": 0.5532176494598389,
+      "eval_runtime": 914.7564,
+      "eval_samples_per_second": 18.759,
+      "eval_steps_per_second": 2.345,
+      "step": 200
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0008666666666666667,
+      "loss": 0.671,
+      "step": 400
+    },
+    {
+      "epoch": 0.82,
+      "eval_exact_match": 0.16095571095571096,
+      "eval_loss": 0.16243064403533936,
+      "eval_runtime": 902.0916,
+      "eval_samples_per_second": 19.022,
+      "eval_steps_per_second": 2.378,
+      "step": 400
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0008,
+      "loss": 0.5276,
+      "step": 600
+    },
+    {
+      "epoch": 1.23,
+      "eval_exact_match": 0.21567599067599066,
+      "eval_loss": 0.06916385143995285,
+      "eval_runtime": 882.9446,
+      "eval_samples_per_second": 19.435,
+      "eval_steps_per_second": 2.429,
+      "step": 600
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0007333333333333333,
+      "loss": 0.4196,
+      "step": 800
+    },
+    {
+      "epoch": 1.64,
+      "eval_exact_match": 0.22587412587412586,
+      "eval_loss": 0.04908030480146408,
+      "eval_runtime": 882.5419,
+      "eval_samples_per_second": 19.444,
+      "eval_steps_per_second": 2.43,
+      "step": 800
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.0006666666666666666,
+      "loss": 0.3593,
+      "step": 1000
+    },
+    {
+      "epoch": 2.05,
+      "eval_exact_match": 0.22913752913752913,
+      "eval_loss": 0.03997402638196945,
+      "eval_runtime": 888.3883,
+      "eval_samples_per_second": 19.316,
+      "eval_steps_per_second": 2.414,
+      "step": 1000
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.0006,
+      "loss": 0.3471,
+      "step": 1200
+    },
+    {
+      "epoch": 2.46,
+      "eval_exact_match": 0.22966200466200465,
+      "eval_loss": 0.03349088877439499,
+      "eval_runtime": 876.9206,
+      "eval_samples_per_second": 19.568,
+      "eval_steps_per_second": 2.446,
+      "step": 1200
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.0005333333333333334,
+      "loss": 0.3416,
+      "step": 1400
+    },
+    {
+      "epoch": 2.87,
+      "eval_exact_match": 0.23175990675990676,
+      "eval_loss": 0.03065803460776806,
+      "eval_runtime": 880.241,
+      "eval_samples_per_second": 19.495,
+      "eval_steps_per_second": 2.437,
+      "step": 1400
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00046666666666666666,
+      "loss": 0.3351,
+      "step": 1600
+    },
+    {
+      "epoch": 3.29,
+      "eval_exact_match": 0.23344988344988346,
+      "eval_loss": 0.030744880437850952,
+      "eval_runtime": 884.2271,
+      "eval_samples_per_second": 19.407,
+      "eval_steps_per_second": 2.426,
+      "step": 1600
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 0.0004,
+      "loss": 0.3316,
+      "step": 1800
+    },
+    {
+      "epoch": 3.7,
+      "eval_exact_match": 0.23432400932400932,
+      "eval_loss": 0.029669322073459625,
+      "eval_runtime": 876.7173,
+      "eval_samples_per_second": 19.573,
+      "eval_steps_per_second": 2.447,
+      "step": 1800
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 0.0003333333333333333,
+      "loss": 0.3312,
+      "step": 2000
+    },
+    {
+      "epoch": 4.11,
+      "eval_exact_match": 0.23444055944055944,
+      "eval_loss": 0.02815121039748192,
+      "eval_runtime": 885.0693,
+      "eval_samples_per_second": 19.388,
+      "eval_steps_per_second": 2.424,
+      "step": 2000
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 0.0002666666666666667,
+      "loss": 0.3271,
+      "step": 2200
+    },
+    {
+      "epoch": 4.52,
+      "eval_exact_match": 0.23648018648018648,
+      "eval_loss": 0.02621879242360592,
+      "eval_runtime": 890.2789,
+      "eval_samples_per_second": 19.275,
+      "eval_steps_per_second": 2.409,
+      "step": 2200
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 0.0002,
+      "loss": 0.3241,
+      "step": 2400
+    },
+    {
+      "epoch": 4.93,
+      "eval_exact_match": 0.23653846153846153,
+      "eval_loss": 0.02629098668694496,
+      "eval_runtime": 876.3961,
+      "eval_samples_per_second": 19.58,
+      "eval_steps_per_second": 2.448,
+      "step": 2400
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.3227,
+      "step": 2600
+    },
+    {
+      "epoch": 5.34,
+      "eval_exact_match": 0.23677156177156178,
+      "eval_loss": 0.02586781419813633,
+      "eval_runtime": 890.1947,
+      "eval_samples_per_second": 19.277,
+      "eval_steps_per_second": 2.41,
+      "step": 2600
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.3201,
+      "step": 2800
+    },
+    {
+      "epoch": 5.75,
+      "eval_exact_match": 0.23653846153846153,
+      "eval_loss": 0.025654641911387444,
+      "eval_runtime": 876.5182,
+      "eval_samples_per_second": 19.577,
+      "eval_steps_per_second": 2.447,
+      "step": 2800
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 0.0,
+      "loss": 0.3227,
+      "step": 3000
+    },
+    {
+      "epoch": 6.16,
+      "eval_exact_match": 0.23653846153846153,
+      "eval_loss": 0.02550993673503399,
+      "eval_runtime": 881.7574,
+      "eval_samples_per_second": 19.461,
+      "eval_steps_per_second": 2.433,
+      "step": 3000
+    },
+    {
+      "epoch": 6.16,
+      "step": 3000,
+      "total_flos": 8.2913691900475e+16,
+      "train_loss": 0.5049722137451171,
+      "train_runtime": 47265.062,
+      "train_samples_per_second": 32.498,
+      "train_steps_per_second": 0.063
+    }
+  ],
+  "max_steps": 3000,
+  "num_train_epochs": 7,
+  "total_flos": 8.2913691900475e+16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4308c72c7ff07ac7aa062d27f37e4fe12d25e31fc93945edbf568fb62fcd8f70
+size 3579