End of training

README.md

@@ -3,11 +3,23 @@ license: llama2
 base_model: meta-llama/Llama-2-7b-hf
 tags:
 - generated_from_trainer
+datasets:
+- tyzhu/lmind_nq_train6000_eval6489_v1_docidx_v3
 metrics:
 - accuracy
 model-index:
 - name: lmind_nq_train6000_eval6489_v1_docidx_v3_meta-llama_Llama-2-7b-hf_lora2
-  results: []
+  results:
+  - task:
+      name: Causal Language Modeling
+      type: text-generation
+    dataset:
+      name: tyzhu/lmind_nq_train6000_eval6489_v1_docidx_v3
+      type: tyzhu/lmind_nq_train6000_eval6489_v1_docidx_v3
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.4470769230769231
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -15,7 +27,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # lmind_nq_train6000_eval6489_v1_docidx_v3_meta-llama_Llama-2-7b-hf_lora2
 
-This model is a fine-tuned version of [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) on the tyzhu/lmind_nq_train6000_eval6489_v1_docidx_v3 dataset.
 It achieves the following results on the evaluation set:
 - Loss: 4.3991
 - Accuracy: 0.4471

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 9.99,
+    "eval_accuracy": 0.4470769230769231,
+    "eval_loss": 4.39909029006958,
+    "eval_runtime": 5.0156,
+    "eval_samples": 500,
+    "eval_samples_per_second": 99.688,
+    "eval_steps_per_second": 12.561,
+    "perplexity": 81.37680569392009,
+    "train_loss": 0.9158074310448163,
+    "train_runtime": 2920.4519,
+    "train_samples": 10925,
+    "train_samples_per_second": 37.409,
+    "train_steps_per_second": 1.168
+}

eval_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 9.99,
+    "eval_accuracy": 0.4470769230769231,
+    "eval_loss": 4.39909029006958,
+    "eval_runtime": 5.0156,
+    "eval_samples": 500,
+    "eval_samples_per_second": 99.688,
+    "eval_steps_per_second": 12.561,
+    "perplexity": 81.37680569392009
+}

tokenizer.json

@@ -1,11 +1,6 @@
 {
   "version": "1.0",
-  "truncation": {
-    "direction": "Right",
-    "max_length": 1024,
-    "strategy": "LongestFirst",
-    "stride": 0
-  },
+  "truncation": null,
   "padding": null,
   "added_tokens": [
     {

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 9.99,
+    "train_loss": 0.9158074310448163,
+    "train_runtime": 2920.4519,
+    "train_samples": 10925,
+    "train_samples_per_second": 37.409,
+    "train_steps_per_second": 1.168
+}

trainer_state.json

@@ -0,0 +1,322 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 9.985358711566619,
+  "eval_steps": 500,
+  "global_step": 3410,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001,
+      "loss": 1.4842,
+      "step": 100
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001,
+      "loss": 1.4016,
+      "step": 200
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001,
+      "loss": 1.3892,
+      "step": 300
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.4543589743589744,
+      "eval_loss": 3.4055869579315186,
+      "eval_runtime": 4.7835,
+      "eval_samples_per_second": 104.525,
+      "eval_steps_per_second": 13.17,
+      "step": 341
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0001,
+      "loss": 1.3608,
+      "step": 400
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0001,
+      "loss": 1.3456,
+      "step": 500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0001,
+      "loss": 1.3499,
+      "step": 600
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.4576923076923077,
+      "eval_loss": 3.453113317489624,
+      "eval_runtime": 4.7768,
+      "eval_samples_per_second": 104.672,
+      "eval_steps_per_second": 13.189,
+      "step": 683
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.0001,
+      "loss": 1.3238,
+      "step": 700
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.0001,
+      "loss": 1.2486,
+      "step": 800
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.0001,
+      "loss": 1.2452,
+      "step": 900
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.0001,
+      "loss": 1.2427,
+      "step": 1000
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.4584102564102564,
+      "eval_loss": 3.671137809753418,
+      "eval_runtime": 5.1059,
+      "eval_samples_per_second": 97.927,
+      "eval_steps_per_second": 12.339,
+      "step": 1024
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.0001,
+      "loss": 1.1507,
+      "step": 1100
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 0.0001,
+      "loss": 1.1218,
+      "step": 1200
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 0.0001,
+      "loss": 1.1231,
+      "step": 1300
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.45697435897435895,
+      "eval_loss": 3.7999510765075684,
+      "eval_runtime": 4.8751,
+      "eval_samples_per_second": 102.561,
+      "eval_steps_per_second": 12.923,
+      "step": 1366
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 0.0001,
+      "loss": 1.0823,
+      "step": 1400
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 0.0001,
+      "loss": 0.997,
+      "step": 1500
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 0.0001,
+      "loss": 1.0024,
+      "step": 1600
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 0.0001,
+      "loss": 0.995,
+      "step": 1700
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.4552307692307692,
+      "eval_loss": 3.953216314315796,
+      "eval_runtime": 4.9741,
+      "eval_samples_per_second": 100.522,
+      "eval_steps_per_second": 12.666,
+      "step": 1707
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 0.0001,
+      "loss": 0.8495,
+      "step": 1800
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 0.0001,
+      "loss": 0.854,
+      "step": 1900
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 0.0001,
+      "loss": 0.8693,
+      "step": 2000
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.45261538461538464,
+      "eval_loss": 4.0766072273254395,
+      "eval_runtime": 5.3625,
+      "eval_samples_per_second": 93.24,
+      "eval_steps_per_second": 11.748,
+      "step": 2049
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 0.0001,
+      "loss": 0.7917,
+      "step": 2100
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 0.0001,
+      "loss": 0.7061,
+      "step": 2200
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 0.0001,
+      "loss": 0.7302,
+      "step": 2300
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.4501025641025641,
+      "eval_loss": 4.171727180480957,
+      "eval_runtime": 5.1508,
+      "eval_samples_per_second": 97.073,
+      "eval_steps_per_second": 12.231,
+      "step": 2390
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 0.0001,
+      "loss": 0.7227,
+      "step": 2400
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 0.0001,
+      "loss": 0.573,
+      "step": 2500
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 0.0001,
+      "loss": 0.6036,
+      "step": 2600
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 0.0001,
+      "loss": 0.6033,
+      "step": 2700
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.448,
+      "eval_loss": 4.277780055999756,
+      "eval_runtime": 4.647,
+      "eval_samples_per_second": 107.596,
+      "eval_steps_per_second": 13.557,
+      "step": 2732
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 0.0001,
+      "loss": 0.4945,
+      "step": 2800
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 0.0001,
+      "loss": 0.4718,
+      "step": 2900
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 0.0001,
+      "loss": 0.4825,
+      "step": 3000
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.4462051282051282,
+      "eval_loss": 4.34151554107666,
+      "eval_runtime": 5.1246,
+      "eval_samples_per_second": 97.568,
+      "eval_steps_per_second": 12.294,
+      "step": 3073
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 0.0001,
+      "loss": 0.4538,
+      "step": 3100
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 0.0001,
+      "loss": 0.3609,
+      "step": 3200
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 0.0001,
+      "loss": 0.3769,
+      "step": 3300
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 0.0001,
+      "loss": 0.387,
+      "step": 3400
+    },
+    {
+      "epoch": 9.99,
+      "eval_accuracy": 0.4470769230769231,
+      "eval_loss": 4.39909029006958,
+      "eval_runtime": 4.6505,
+      "eval_samples_per_second": 107.514,
+      "eval_steps_per_second": 13.547,
+      "step": 3410
+    },
+    {
+      "epoch": 9.99,
+      "step": 3410,
+      "total_flos": 6.753170451008061e+17,
+      "train_loss": 0.9158074310448163,
+      "train_runtime": 2920.4519,
+      "train_samples_per_second": 37.409,
+      "train_steps_per_second": 1.168
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 3410,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "total_flos": 6.753170451008061e+17,
+  "trial_name": null,
+  "trial_params": null
+}