End of training

README.md

@@ -3,6 +3,7 @@ license: bigscience-bloom-rail-1.0
 library_name: peft
 tags:
 - llama-factory
+- lora
 - generated_from_trainer
 base_model: bigscience/bloom-7b1
 model-index:
@@ -15,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # BLOOM_AAID_new_mixed_train_final
 
-This model is a fine-tuned version of [bigscience/bloom-7b1](https://huggingface.co/bigscience/bloom-7b1) on an unknown dataset.
+This model is a fine-tuned version of [bigscience/bloom-7b1](https://huggingface.co/bigscience/bloom-7b1) on the AAID_new_mixed dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8549
+- Loss: 0.8218
 
 ## Model description
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 0.3390757451462948,
+    "eval_loss": 0.8218151926994324,
+    "eval_runtime": 209.1208,
+    "eval_samples_per_second": 22.532,
+    "eval_steps_per_second": 0.708,
+    "total_flos": 4.228376200709407e+17,
+    "train_loss": 0.4842093152384604,
+    "train_runtime": 19786.6932,
+    "train_samples_per_second": 35.484,
+    "train_steps_per_second": 0.139
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 0.3390757451462948,
+    "eval_loss": 0.8218151926994324,
+    "eval_runtime": 209.1208,
+    "eval_samples_per_second": 22.532,
+    "eval_steps_per_second": 0.708
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 0.3390757451462948,
+    "total_flos": 4.228376200709407e+17,
+    "train_loss": 0.4842093152384604,
+    "train_runtime": 19786.6932,
+    "train_samples_per_second": 35.484,
+    "train_steps_per_second": 0.139
+}

trainer_state.json

@@ -0,0 +1,516 @@
+{
+  "best_metric": 0.8218151926994324,
+  "best_model_checkpoint": "saves/BLOOM-7B/lora/train_1/checkpoint-210",
+  "epoch": 0.3390757451462948,
+  "eval_steps": 10,
+  "global_step": 310,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.010937927262783703,
+      "grad_norm": 1.1790575981140137,
+      "learning_rate": 0.00029999015487222375,
+      "loss": 1.6811,
+      "step": 10
+    },
+    {
+      "epoch": 0.010937927262783703,
+      "eval_loss": 0.9062243103981018,
+      "eval_runtime": 210.246,
+      "eval_samples_per_second": 22.412,
+      "eval_steps_per_second": 0.704,
+      "step": 10
+    },
+    {
+      "epoch": 0.021875854525567406,
+      "grad_norm": 0.5708920359611511,
+      "learning_rate": 0.00029996062078124905,
+      "loss": 0.6248,
+      "step": 20
+    },
+    {
+      "epoch": 0.021875854525567406,
+      "eval_loss": 0.8811991810798645,
+      "eval_runtime": 210.0954,
+      "eval_samples_per_second": 22.428,
+      "eval_steps_per_second": 0.704,
+      "step": 20
+    },
+    {
+      "epoch": 0.03281378178835111,
+      "grad_norm": 0.5245503783226013,
+      "learning_rate": 0.0002999114016039678,
+      "loss": 0.5466,
+      "step": 30
+    },
+    {
+      "epoch": 0.03281378178835111,
+      "eval_loss": 0.8906806707382202,
+      "eval_runtime": 210.0232,
+      "eval_samples_per_second": 22.436,
+      "eval_steps_per_second": 0.705,
+      "step": 30
+    },
+    {
+      "epoch": 0.04375170905113481,
+      "grad_norm": 0.31997227668762207,
+      "learning_rate": 0.00029984250380130117,
+      "loss": 0.5591,
+      "step": 40
+    },
+    {
+      "epoch": 0.04375170905113481,
+      "eval_loss": 0.8906031250953674,
+      "eval_runtime": 209.6861,
+      "eval_samples_per_second": 22.472,
+      "eval_steps_per_second": 0.706,
+      "step": 40
+    },
+    {
+      "epoch": 0.05468963631391851,
+      "grad_norm": 0.2958827614784241,
+      "learning_rate": 0.0002997539364173515,
+      "loss": 0.5318,
+      "step": 50
+    },
+    {
+      "epoch": 0.05468963631391851,
+      "eval_loss": 0.8625577688217163,
+      "eval_runtime": 210.182,
+      "eval_samples_per_second": 22.419,
+      "eval_steps_per_second": 0.704,
+      "step": 50
+    },
+    {
+      "epoch": 0.06562756357670221,
+      "grad_norm": 0.396158367395401,
+      "learning_rate": 0.00029964571107821494,
+      "loss": 0.496,
+      "step": 60
+    },
+    {
+      "epoch": 0.06562756357670221,
+      "eval_loss": 0.8501101136207581,
+      "eval_runtime": 210.0313,
+      "eval_samples_per_second": 22.435,
+      "eval_steps_per_second": 0.705,
+      "step": 60
+    },
+    {
+      "epoch": 0.07656549083948591,
+      "grad_norm": 0.4420228600502014,
+      "learning_rate": 0.00029951784199045534,
+      "loss": 0.4855,
+      "step": 70
+    },
+    {
+      "epoch": 0.07656549083948591,
+      "eval_loss": 0.8276830911636353,
+      "eval_runtime": 210.1084,
+      "eval_samples_per_second": 22.427,
+      "eval_steps_per_second": 0.704,
+      "step": 70
+    },
+    {
+      "epoch": 0.08750341810226962,
+      "grad_norm": 0.4005095958709717,
+      "learning_rate": 0.0002993703459392396,
+      "loss": 0.4746,
+      "step": 80
+    },
+    {
+      "epoch": 0.08750341810226962,
+      "eval_loss": 0.8740801215171814,
+      "eval_runtime": 210.097,
+      "eval_samples_per_second": 22.428,
+      "eval_steps_per_second": 0.704,
+      "step": 80
+    },
+    {
+      "epoch": 0.09844134536505332,
+      "grad_norm": 0.4457601308822632,
+      "learning_rate": 0.00029920324228613376,
+      "loss": 0.4846,
+      "step": 90
+    },
+    {
+      "epoch": 0.09844134536505332,
+      "eval_loss": 0.8443948030471802,
+      "eval_runtime": 209.5315,
+      "eval_samples_per_second": 22.488,
+      "eval_steps_per_second": 0.706,
+      "step": 90
+    },
+    {
+      "epoch": 0.10937927262783702,
+      "grad_norm": 0.3862694799900055,
+      "learning_rate": 0.0002990165529665622,
+      "loss": 0.4424,
+      "step": 100
+    },
+    {
+      "epoch": 0.10937927262783702,
+      "eval_loss": 0.843928873538971,
+      "eval_runtime": 210.1339,
+      "eval_samples_per_second": 22.424,
+      "eval_steps_per_second": 0.704,
+      "step": 100
+    },
+    {
+      "epoch": 0.12031719989062073,
+      "grad_norm": 0.32919445633888245,
+      "learning_rate": 0.0002988103024869277,
+      "loss": 0.4514,
+      "step": 110
+    },
+    {
+      "epoch": 0.12031719989062073,
+      "eval_loss": 0.8323902487754822,
+      "eval_runtime": 210.1443,
+      "eval_samples_per_second": 22.423,
+      "eval_steps_per_second": 0.704,
+      "step": 110
+    },
+    {
+      "epoch": 0.13125512715340443,
+      "grad_norm": 0.3766002953052521,
+      "learning_rate": 0.00029858451792139453,
+      "loss": 0.4485,
+      "step": 120
+    },
+    {
+      "epoch": 0.13125512715340443,
+      "eval_loss": 0.8639554381370544,
+      "eval_runtime": 210.1293,
+      "eval_samples_per_second": 22.424,
+      "eval_steps_per_second": 0.704,
+      "step": 120
+    },
+    {
+      "epoch": 0.14219305441618812,
+      "grad_norm": 0.3994982838630676,
+      "learning_rate": 0.0002983392289083346,
+      "loss": 0.443,
+      "step": 130
+    },
+    {
+      "epoch": 0.14219305441618812,
+      "eval_loss": 0.853752076625824,
+      "eval_runtime": 210.1289,
+      "eval_samples_per_second": 22.424,
+      "eval_steps_per_second": 0.704,
+      "step": 130
+    },
+    {
+      "epoch": 0.15313098167897182,
+      "grad_norm": 0.2923962473869324,
+      "learning_rate": 0.0002980744676464371,
+      "loss": 0.4316,
+      "step": 140
+    },
+    {
+      "epoch": 0.15313098167897182,
+      "eval_loss": 0.826151430606842,
+      "eval_runtime": 209.9359,
+      "eval_samples_per_second": 22.445,
+      "eval_steps_per_second": 0.705,
+      "step": 140
+    },
+    {
+      "epoch": 0.16406890894175555,
+      "grad_norm": 0.33790677785873413,
+      "learning_rate": 0.0002977902688904813,
+      "loss": 0.4435,
+      "step": 150
+    },
+    {
+      "epoch": 0.16406890894175555,
+      "eval_loss": 0.855368971824646,
+      "eval_runtime": 210.0755,
+      "eval_samples_per_second": 22.43,
+      "eval_steps_per_second": 0.705,
+      "step": 150
+    },
+    {
+      "epoch": 0.17500683620453925,
+      "grad_norm": 0.3131038248538971,
+      "learning_rate": 0.00029748666994677467,
+      "loss": 0.4269,
+      "step": 160
+    },
+    {
+      "epoch": 0.17500683620453925,
+      "eval_loss": 0.8361687660217285,
+      "eval_runtime": 209.5838,
+      "eval_samples_per_second": 22.483,
+      "eval_steps_per_second": 0.706,
+      "step": 160
+    },
+    {
+      "epoch": 0.18594476346732294,
+      "grad_norm": 0.35536178946495056,
+      "learning_rate": 0.00029716371066825593,
+      "loss": 0.4322,
+      "step": 170
+    },
+    {
+      "epoch": 0.18594476346732294,
+      "eval_loss": 0.8331068158149719,
+      "eval_runtime": 210.059,
+      "eval_samples_per_second": 22.432,
+      "eval_steps_per_second": 0.705,
+      "step": 170
+    },
+    {
+      "epoch": 0.19688269073010664,
+      "grad_norm": 0.3734683692455292,
+      "learning_rate": 0.0002968214334492632,
+      "loss": 0.4136,
+      "step": 180
+    },
+    {
+      "epoch": 0.19688269073010664,
+      "eval_loss": 0.8500174283981323,
+      "eval_runtime": 209.7577,
+      "eval_samples_per_second": 22.464,
+      "eval_steps_per_second": 0.706,
+      "step": 180
+    },
+    {
+      "epoch": 0.20782061799289034,
+      "grad_norm": 0.3501605987548828,
+      "learning_rate": 0.00029645988321996917,
+      "loss": 0.4262,
+      "step": 190
+    },
+    {
+      "epoch": 0.20782061799289034,
+      "eval_loss": 0.8315255641937256,
+      "eval_runtime": 209.8869,
+      "eval_samples_per_second": 22.45,
+      "eval_steps_per_second": 0.705,
+      "step": 190
+    },
+    {
+      "epoch": 0.21875854525567404,
+      "grad_norm": 0.2926044464111328,
+      "learning_rate": 0.00029607910744048336,
+      "loss": 0.4283,
+      "step": 200
+    },
+    {
+      "epoch": 0.21875854525567404,
+      "eval_loss": 0.8357769846916199,
+      "eval_runtime": 210.1683,
+      "eval_samples_per_second": 22.42,
+      "eval_steps_per_second": 0.704,
+      "step": 200
+    },
+    {
+      "epoch": 0.22969647251845776,
+      "grad_norm": 0.4317471385002136,
+      "learning_rate": 0.00029567915609462174,
+      "loss": 0.3983,
+      "step": 210
+    },
+    {
+      "epoch": 0.22969647251845776,
+      "eval_loss": 0.8218151926994324,
+      "eval_runtime": 210.2221,
+      "eval_samples_per_second": 22.414,
+      "eval_steps_per_second": 0.704,
+      "step": 210
+    },
+    {
+      "epoch": 0.24063439978124146,
+      "grad_norm": 0.4090133309364319,
+      "learning_rate": 0.00029526008168334573,
+      "loss": 0.409,
+      "step": 220
+    },
+    {
+      "epoch": 0.24063439978124146,
+      "eval_loss": 0.8271812200546265,
+      "eval_runtime": 209.5484,
+      "eval_samples_per_second": 22.486,
+      "eval_steps_per_second": 0.706,
+      "step": 220
+    },
+    {
+      "epoch": 0.25157232704402516,
+      "grad_norm": 0.46969935297966003,
+      "learning_rate": 0.0002948219392178703,
+      "loss": 0.4173,
+      "step": 230
+    },
+    {
+      "epoch": 0.25157232704402516,
+      "eval_loss": 0.8557965159416199,
+      "eval_runtime": 210.1229,
+      "eval_samples_per_second": 22.425,
+      "eval_steps_per_second": 0.704,
+      "step": 230
+    },
+    {
+      "epoch": 0.26251025430680885,
+      "grad_norm": 0.3677867650985718,
+      "learning_rate": 0.0002943647862124429,
+      "loss": 0.3847,
+      "step": 240
+    },
+    {
+      "epoch": 0.26251025430680885,
+      "eval_loss": 0.8389946818351746,
+      "eval_runtime": 210.1338,
+      "eval_samples_per_second": 22.424,
+      "eval_steps_per_second": 0.704,
+      "step": 240
+    },
+    {
+      "epoch": 0.27344818156959255,
+      "grad_norm": 0.43112123012542725,
+      "learning_rate": 0.0002938886826767936,
+      "loss": 0.3923,
+      "step": 250
+    },
+    {
+      "epoch": 0.27344818156959255,
+      "eval_loss": 0.8553555011749268,
+      "eval_runtime": 209.3302,
+      "eval_samples_per_second": 22.51,
+      "eval_steps_per_second": 0.707,
+      "step": 250
+    },
+    {
+      "epoch": 0.28438610883237625,
+      "grad_norm": 0.4160371422767639,
+      "learning_rate": 0.00029339369110825756,
+      "loss": 0.4014,
+      "step": 260
+    },
+    {
+      "epoch": 0.28438610883237625,
+      "eval_loss": 0.8533877730369568,
+      "eval_runtime": 209.8359,
+      "eval_samples_per_second": 22.456,
+      "eval_steps_per_second": 0.705,
+      "step": 260
+    },
+    {
+      "epoch": 0.29532403609515995,
+      "grad_norm": 0.426897794008255,
+      "learning_rate": 0.00029287987648357134,
+      "loss": 0.3967,
+      "step": 270
+    },
+    {
+      "epoch": 0.29532403609515995,
+      "eval_loss": 0.8450255990028381,
+      "eval_runtime": 209.7931,
+      "eval_samples_per_second": 22.46,
+      "eval_steps_per_second": 0.705,
+      "step": 270
+    },
+    {
+      "epoch": 0.30626196335794365,
+      "grad_norm": 0.42360490560531616,
+      "learning_rate": 0.00029234730625034343,
+      "loss": 0.3936,
+      "step": 280
+    },
+    {
+      "epoch": 0.30626196335794365,
+      "eval_loss": 0.8424352407455444,
+      "eval_runtime": 209.6462,
+      "eval_samples_per_second": 22.476,
+      "eval_steps_per_second": 0.706,
+      "step": 280
+    },
+    {
+      "epoch": 0.3171998906207274,
+      "grad_norm": 0.3254799246788025,
+      "learning_rate": 0.00029179605031820044,
+      "loss": 0.3816,
+      "step": 290
+    },
+    {
+      "epoch": 0.3171998906207274,
+      "eval_loss": 0.8733253479003906,
+      "eval_runtime": 210.002,
+      "eval_samples_per_second": 22.438,
+      "eval_steps_per_second": 0.705,
+      "step": 290
+    },
+    {
+      "epoch": 0.3281378178835111,
+      "grad_norm": 0.44433578848838806,
+      "learning_rate": 0.0002912261810496102,
+      "loss": 0.3852,
+      "step": 300
+    },
+    {
+      "epoch": 0.3281378178835111,
+      "eval_loss": 0.8514276146888733,
+      "eval_runtime": 209.6715,
+      "eval_samples_per_second": 22.473,
+      "eval_steps_per_second": 0.706,
+      "step": 300
+    },
+    {
+      "epoch": 0.3390757451462948,
+      "grad_norm": 0.3501994013786316,
+      "learning_rate": 0.0002906377732503829,
+      "loss": 0.379,
+      "step": 310
+    },
+    {
+      "epoch": 0.3390757451462948,
+      "eval_loss": 0.854888379573822,
+      "eval_runtime": 210.2495,
+      "eval_samples_per_second": 22.411,
+      "eval_steps_per_second": 0.704,
+      "step": 310
+    },
+    {
+      "epoch": 0.3390757451462948,
+      "step": 310,
+      "total_flos": 4.228376200709407e+17,
+      "train_loss": 0.4842093152384604,
+      "train_runtime": 19786.6932,
+      "train_samples_per_second": 35.484,
+      "train_steps_per_second": 0.139
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2742,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 10,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 10,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.228376200709407e+17,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}