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_2
 
-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: 0.8901
+- Loss: 0.8316
 
 ## Model description
 

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 0.3937653814602133,
+    "eval_loss": 0.83160400390625,
+    "eval_runtime": 209.9462,
+    "eval_samples_per_second": 22.444,
+    "eval_steps_per_second": 0.705,
+    "total_flos": 4.90773729509376e+17,
+    "train_loss": 0.6297279569837783,
+    "train_runtime": 17337.0867,
+    "train_samples_per_second": 40.498,
+    "train_steps_per_second": 0.158
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 0.3937653814602133,
+    "eval_loss": 0.83160400390625,
+    "eval_runtime": 209.9462,
+    "eval_samples_per_second": 22.444,
+    "eval_steps_per_second": 0.705
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 0.3937653814602133,
+    "total_flos": 4.90773729509376e+17,
+    "train_loss": 0.6297279569837783,
+    "train_runtime": 17337.0867,
+    "train_samples_per_second": 40.498,
+    "train_steps_per_second": 0.158
+}

trainer_state.json

@@ -0,0 +1,186 @@
+{
+  "best_metric": 0.83160400390625,
+  "best_model_checkpoint": "saves/BLOOM-7B/lora/train_1/checkpoint-160",
+  "epoch": 0.3937653814602133,
+  "eval_steps": 40,
+  "global_step": 360,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.04375170905113481,
+      "grad_norm": 0.503142237663269,
+      "learning_rate": 0.00014457831325301204,
+      "loss": 2.1275,
+      "step": 40
+    },
+    {
+      "epoch": 0.04375170905113481,
+      "eval_loss": 0.8730847239494324,
+      "eval_runtime": 210.0237,
+      "eval_samples_per_second": 22.436,
+      "eval_steps_per_second": 0.705,
+      "step": 40
+    },
+    {
+      "epoch": 0.08750341810226962,
+      "grad_norm": 0.5874570608139038,
+      "learning_rate": 0.0002891566265060241,
+      "loss": 0.5607,
+      "step": 80
+    },
+    {
+      "epoch": 0.08750341810226962,
+      "eval_loss": 0.9147108197212219,
+      "eval_runtime": 210.6077,
+      "eval_samples_per_second": 22.373,
+      "eval_steps_per_second": 0.703,
+      "step": 80
+    },
+    {
+      "epoch": 0.13125512715340443,
+      "grad_norm": 0.6330702900886536,
+      "learning_rate": 0.00029985669589905196,
+      "loss": 0.4772,
+      "step": 120
+    },
+    {
+      "epoch": 0.13125512715340443,
+      "eval_loss": 0.8467113971710205,
+      "eval_runtime": 210.5882,
+      "eval_samples_per_second": 22.375,
+      "eval_steps_per_second": 0.703,
+      "step": 120
+    },
+    {
+      "epoch": 0.17500683620453925,
+      "grad_norm": 0.3914893865585327,
+      "learning_rate": 0.0002993796936474492,
+      "loss": 0.4536,
+      "step": 160
+    },
+    {
+      "epoch": 0.17500683620453925,
+      "eval_loss": 0.83160400390625,
+      "eval_runtime": 210.0873,
+      "eval_samples_per_second": 22.429,
+      "eval_steps_per_second": 0.704,
+      "step": 160
+    },
+    {
+      "epoch": 0.21875854525567404,
+      "grad_norm": 0.3717311918735504,
+      "learning_rate": 0.00029856911617379416,
+      "loss": 0.4388,
+      "step": 200
+    },
+    {
+      "epoch": 0.21875854525567404,
+      "eval_loss": 0.8363510370254517,
+      "eval_runtime": 210.5622,
+      "eval_samples_per_second": 22.378,
+      "eval_steps_per_second": 0.703,
+      "step": 200
+    },
+    {
+      "epoch": 0.26251025430680885,
+      "grad_norm": 0.3580843508243561,
+      "learning_rate": 0.00029742677355383513,
+      "loss": 0.4156,
+      "step": 240
+    },
+    {
+      "epoch": 0.26251025430680885,
+      "eval_loss": 0.8319957852363586,
+      "eval_runtime": 210.5767,
+      "eval_samples_per_second": 22.377,
+      "eval_steps_per_second": 0.703,
+      "step": 240
+    },
+    {
+      "epoch": 0.30626196335794365,
+      "grad_norm": 0.43683090806007385,
+      "learning_rate": 0.0002959552167179149,
+      "loss": 0.4083,
+      "step": 280
+    },
+    {
+      "epoch": 0.30626196335794365,
+      "eval_loss": 0.859396755695343,
+      "eval_runtime": 210.6094,
+      "eval_samples_per_second": 22.373,
+      "eval_steps_per_second": 0.703,
+      "step": 280
+    },
+    {
+      "epoch": 0.3500136724090785,
+      "grad_norm": 0.3830896317958832,
+      "learning_rate": 0.00029415773175456614,
+      "loss": 0.397,
+      "step": 320
+    },
+    {
+      "epoch": 0.3500136724090785,
+      "eval_loss": 0.859840989112854,
+      "eval_runtime": 210.7225,
+      "eval_samples_per_second": 22.361,
+      "eval_steps_per_second": 0.702,
+      "step": 320
+    },
+    {
+      "epoch": 0.3937653814602133,
+      "grad_norm": 0.5180444717407227,
+      "learning_rate": 0.0002920383325724476,
+      "loss": 0.3887,
+      "step": 360
+    },
+    {
+      "epoch": 0.3937653814602133,
+      "eval_loss": 0.890067458152771,
+      "eval_runtime": 210.5277,
+      "eval_samples_per_second": 22.382,
+      "eval_steps_per_second": 0.703,
+      "step": 360
+    },
+    {
+      "epoch": 0.3937653814602133,
+      "step": 360,
+      "total_flos": 4.90773729509376e+17,
+      "train_loss": 0.6297279569837783,
+      "train_runtime": 17337.0867,
+      "train_samples_per_second": 40.498,
+      "train_steps_per_second": 0.158
+    }
+  ],
+  "logging_steps": 40,
+  "max_steps": 2742,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 40,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "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": 4.90773729509376e+17,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}