commit files to HF hub

README.md

@@ -2,19 +2,32 @@
 license: apache-2.0
 tags:
 - generated_from_trainer
+datasets:
+- Multi-Domain-Expert-Layers/pubmed_central
 metrics:
 - accuracy
 model-index:
-- name: expert-pubmed_central
-  results: []
+- name: layer_9,10,11,12,13
+  results:
+  - task:
+      type: text-generation
+      name: Causal Language Modeling
+    dataset:
+      name: Multi-Domain-Expert-Layers/pubmed_central
+      type: Multi-Domain-Expert-Layers/pubmed_central
+      split: None
+    metrics:
+    - type: accuracy
+      value: 0.5767534246575342
+      name: Accuracy
 ---
 
 <!-- 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. -->
 
-# expert-pubmed_central
+# layer_9,10,11,12,13
 
-This model is a fine-tuned version of [EleutherAI/pythia-1b-deduped](https://huggingface.co/EleutherAI/pythia-1b-deduped) on the None dataset.
+This model is a fine-tuned version of [EleutherAI/pythia-1b-deduped](https://huggingface.co/EleutherAI/pythia-1b-deduped) on the Multi-Domain-Expert-Layers/pubmed_central dataset.
 It achieves the following results on the evaluation set:
 - Loss: 2.0227
 - Accuracy: 0.5768
@@ -66,3 +79,7 @@ The following hyperparameters were used during training:
 - Pytorch 2.0.0+cu117
 - Datasets 2.11.0
 - Tokenizers 0.13.3
+
+
+## Wandb Report
+ https://wandb.ai/ontocord/pythia-1b-deduped-layer-test-pubmed_central/runs/yy3pwx0o

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 0.02,
+    "eval_accuracy": 0.5767534246575342,
+    "eval_loss": 2.0226891040802,
+    "eval_runtime": 13.0624,
+    "eval_samples": 1000,
+    "eval_samples_per_second": 76.555,
+    "eval_steps_per_second": 1.225,
+    "perplexity": 7.558623554318639,
+    "train_loss": 2.0490207977294923,
+    "train_runtime": 1425.429,
+    "train_samples": 3358758,
+    "train_samples_per_second": 44.899,
+    "train_steps_per_second": 0.702
+}

eval_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 0.02,
+    "eval_accuracy": 0.5767534246575342,
+    "eval_loss": 2.0226891040802,
+    "eval_runtime": 13.0624,
+    "eval_samples": 1000,
+    "eval_samples_per_second": 76.555,
+    "eval_steps_per_second": 1.225,
+    "perplexity": 7.558623554318639
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 0.02,
+    "train_loss": 2.0490207977294923,
+    "train_runtime": 1425.429,
+    "train_samples": 3358758,
+    "train_samples_per_second": 44.899,
+    "train_steps_per_second": 0.702
+}

trainer_state.json

@@ -0,0 +1,370 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.019054651121247126,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.8e-05,
+      "loss": 2.0665,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.6e-05,
+      "loss": 2.038,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.4e-05,
+      "loss": 2.0553,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 2.063,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9e-05,
+      "loss": 2.0761,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 2.0427,
+      "step": 120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.6e-05,
+      "loss": 2.0504,
+      "step": 140
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.4e-05,
+      "loss": 2.072,
+      "step": 160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.2e-05,
+      "loss": 2.0897,
+      "step": 180
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8e-05,
+      "loss": 2.0567,
+      "step": 200
+    },
+    {
+      "epoch": 0.0,
+      "eval_accuracy": 0.57173385518591,
+      "eval_loss": 2.053316354751587,
+      "eval_runtime": 13.3251,
+      "eval_samples_per_second": 75.046,
+      "eval_steps_per_second": 1.201,
+      "step": 200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 2.0786,
+      "step": 220
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.6e-05,
+      "loss": 2.0559,
+      "step": 240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.4e-05,
+      "loss": 2.0664,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.2e-05,
+      "loss": 2.0799,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7e-05,
+      "loss": 2.0744,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 2.0509,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.6e-05,
+      "loss": 2.094,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 2.0419,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.2e-05,
+      "loss": 2.055,
+      "step": 380
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6e-05,
+      "loss": 2.041,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "eval_accuracy": 0.5733287671232876,
+      "eval_loss": 2.043827772140503,
+      "eval_runtime": 13.0102,
+      "eval_samples_per_second": 76.863,
+      "eval_steps_per_second": 1.23,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.8e-05,
+      "loss": 2.0314,
+      "step": 420
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 2.0513,
+      "step": 440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 2.0545,
+      "step": 460
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 2.0954,
+      "step": 480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 2.0493,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8e-05,
+      "loss": 2.0714,
+      "step": 520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 2.077,
+      "step": 540
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 2.07,
+      "step": 560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2e-05,
+      "loss": 2.0233,
+      "step": 580
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4e-05,
+      "loss": 2.0496,
+      "step": 600
+    },
+    {
+      "epoch": 0.01,
+      "eval_accuracy": 0.5749158512720156,
+      "eval_loss": 2.036094903945923,
+      "eval_runtime": 12.9978,
+      "eval_samples_per_second": 76.936,
+      "eval_steps_per_second": 1.231,
+      "step": 600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8e-05,
+      "loss": 2.063,
+      "step": 620
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6e-05,
+      "loss": 2.0683,
+      "step": 640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 2.0344,
+      "step": 660
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 2.0414,
+      "step": 680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-05,
+      "loss": 2.0146,
+      "step": 700
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 2.0369,
+      "step": 720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 2.0333,
+      "step": 740
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4e-05,
+      "loss": 2.039,
+      "step": 760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 2.0402,
+      "step": 780
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.0194,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "eval_accuracy": 0.5760665362035226,
+      "eval_loss": 2.0276312828063965,
+      "eval_runtime": 11.977,
+      "eval_samples_per_second": 83.493,
+      "eval_steps_per_second": 1.336,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8e-05,
+      "loss": 2.003,
+      "step": 820
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.0207,
+      "step": 840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 2.0228,
+      "step": 860
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2e-05,
+      "loss": 2.0349,
+      "step": 880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 2.0158,
+      "step": 900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.019,
+      "step": 920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6e-06,
+      "loss": 2.0196,
+      "step": 940
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.0414,
+      "step": 960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.0279,
+      "step": 980
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0,
+      "loss": 2.0338,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02,
+      "eval_accuracy": 0.5767534246575342,
+      "eval_loss": 2.0226891040802,
+      "eval_runtime": 13.0341,
+      "eval_samples_per_second": 76.722,
+      "eval_steps_per_second": 1.228,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02,
+      "step": 1000,
+      "total_flos": 1.7866929733632e+17,
+      "train_loss": 2.0490207977294923,
+      "train_runtime": 1425.429,
+      "train_samples_per_second": 44.899,
+      "train_steps_per_second": 0.702
+    }
+  ],
+  "max_steps": 1000,
+  "num_train_epochs": 1,
+  "total_flos": 1.7866929733632e+17,
+  "trial_name": null,
+  "trial_params": null
+}