Training in progress, step 150, checkpoint

checkpoint-150/README.md

@@ -216,23 +216,4 @@ The following `bitsandbytes` quantization config was used during training:
 ### Framework versions
 
 
-- PEFT 0.6.0
-## Training procedure
-
-
-The following `bitsandbytes` quantization config was used during training:
-- quant_method: bitsandbytes
-- load_in_8bit: True
-- load_in_4bit: False
-- llm_int8_threshold: 6.0
-- llm_int8_skip_modules: None
-- llm_int8_enable_fp32_cpu_offload: False
-- llm_int8_has_fp16_weight: False
-- bnb_4bit_quant_type: fp4
-- bnb_4bit_use_double_quant: False
-- bnb_4bit_compute_dtype: float32
-
-### Framework versions
-
-
 - PEFT 0.6.0

checkpoint-150/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dddcd65ff6df9d9f677eb4f14a43db2eb07482d82b13c30695cff8601f54c15e
+oid sha256:ea50990e2fbe8ad8a5a88316b8fb048d0116c942d21f488d5ccb818a1b8e9e15
 size 19690328

checkpoint-150/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2bf8a8e294ad91dad98cdc10260e4bc5df16015bdea464a7c37a94878393d035
+oid sha256:82cef63642550ff002fae618efdfb668c63baaf5ea586293197a8723e05dee6b
 size 38087098

checkpoint-150/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:55e71ed3f6f3ad40fcbc3c5cfb0b3bc77f998518af052b57e64a7852561eecf3
+oid sha256:0ea9739befcd5c51848e79539ca055a35fe66fb199ad3370f87457aafbf66891
 size 14244

checkpoint-150/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c5dfe32eefef4e0fa34f6f96b3b2ec5e3e27c4f5298718cbc335d8fcca2a95dc
+oid sha256:d7daf604dcf714fec2b8a52977c6234cee0542257c2bce27fffd192279b9364f
 size 1064

checkpoint-150/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 1.2631328105926514,
+  "best_metric": 0.45803794264793396,
   "best_model_checkpoint": "output/checkpoint-150",
-  "epoch": 0.7657945118059988,
+  "epoch": 0.12032487716835456,
   "eval_steps": 50,
   "global_step": 150,
   "is_hyper_param_search": false,
@@ -9,62 +9,62 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.01,
-      "learning_rate": 2.777777777777778e-07,
-      "loss": 9.5569,
+      "epoch": 0.0,
+      "learning_rate": 0.0,
+      "loss": 1.6453,
       "step": 1
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 1.25e-05,
-      "loss": 6.958,
+      "epoch": 0.04,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 1.3216,
       "step": 50
     },
     {
-      "epoch": 0.26,
-      "eval_accuracy": 0.42857142857142855,
-      "eval_loss": 6.581231117248535,
-      "eval_runtime": 28.2966,
-      "eval_samples_per_second": 6.184,
-      "eval_steps_per_second": 1.555,
+      "epoch": 0.04,
+      "eval_accuracy": 0.8057142857142857,
+      "eval_loss": 0.6264824271202087,
+      "eval_runtime": 28.8159,
+      "eval_samples_per_second": 6.073,
+      "eval_steps_per_second": 1.527,
       "step": 50
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 2.6388888888888892e-05,
-      "loss": 1.6141,
+      "epoch": 0.08,
+      "learning_rate": 2.088888888888889e-05,
+      "loss": 0.918,
       "step": 100
     },
     {
-      "epoch": 0.51,
-      "eval_accuracy": 0.5028571428571429,
-      "eval_loss": 1.6640697717666626,
-      "eval_runtime": 28.1139,
-      "eval_samples_per_second": 6.225,
-      "eval_steps_per_second": 1.565,
+      "epoch": 0.08,
+      "eval_accuracy": 0.7885714285714286,
+      "eval_loss": 0.49068325757980347,
+      "eval_runtime": 26.1642,
+      "eval_samples_per_second": 6.689,
+      "eval_steps_per_second": 1.682,
       "step": 100
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.027777777777778e-05,
-      "loss": 0.8533,
+      "epoch": 0.12,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.7219,
       "step": 150
     },
     {
-      "epoch": 0.77,
-      "eval_accuracy": 0.6342857142857142,
-      "eval_loss": 1.2631328105926514,
-      "eval_runtime": 26.205,
-      "eval_samples_per_second": 6.678,
-      "eval_steps_per_second": 1.679,
+      "epoch": 0.12,
+      "eval_accuracy": 0.7942857142857143,
+      "eval_loss": 0.45803794264793396,
+      "eval_runtime": 26.1308,
+      "eval_samples_per_second": 6.697,
+      "eval_steps_per_second": 1.684,
       "step": 150
     }
   ],
   "logging_steps": 50,
-  "max_steps": 1000,
-  "num_train_epochs": 6,
+  "max_steps": 1246,
+  "num_train_epochs": 1,
   "save_steps": 50,
-  "total_flos": 5.595400201863168e+16,
+  "total_flos": 5.557605717368832e+16,
   "trial_name": null,
   "trial_params": null
 }

checkpoint-150/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5a0ef36d7c31265f68a3edc8ae53ce9272dcc90a79f261e29a2a46e040c77a33
+oid sha256:2b7c1677f66b49d74d914a320393e2fa29530a9670d5824a30beac9590468d16
 size 4600