Training in progress, epoch 6, checkpoint

checkpoint-1272/adapter_config.json

@@ -14,7 +14,7 @@
     "classifier.weight"
   ],
   "peft_type": "LORA",
-  "r": 128,
+  "r": 64,
   "revision": null,
   "target_modules": [
     "query",

checkpoint-1272/adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9d0fcd50de5b29a0b81ac6a1b131f337af435d4794b2f2322d357fda68fdad05
-size 230550429
+oid sha256:0f35c03d7a318d568526199e88bda759a33e71b33dbfffb4d6fd3fbaaa2a8276
+size 117304221

checkpoint-1272/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f83d9598b7487d67e421285999323e5e383bf6a5e7afa1e08ccba12565ecaea8
-size 461144966
+oid sha256:e57755f3bcba71f91fb6a37ff2e2e719cbd33a52b40beed58a83a51175028b18
+size 234652550

checkpoint-1272/trainer_state.json

@@ -11,85 +11,85 @@
     {
       "epoch": 1.0,
       "learning_rate": 9.95e-05,
-      "loss": 6.3737,
+      "loss": 6.3714,
       "step": 212
     },
     {
       "epoch": 1.0,
-      "eval_loss": 5.842774868011475,
-      "eval_runtime": 3.9529,
-      "eval_samples_per_second": 171.015,
-      "eval_steps_per_second": 21.503,
+      "eval_loss": 5.850927352905273,
+      "eval_runtime": 3.8282,
+      "eval_samples_per_second": 176.584,
+      "eval_steps_per_second": 22.204,
       "step": 212
     },
     {
       "epoch": 2.0,
       "learning_rate": 9.900000000000001e-05,
-      "loss": 5.5296,
+      "loss": 5.5438,
       "step": 424
     },
     {
       "epoch": 2.0,
-      "eval_loss": 4.859792232513428,
-      "eval_runtime": 4.2848,
-      "eval_samples_per_second": 157.766,
-      "eval_steps_per_second": 19.837,
+      "eval_loss": 4.862037658691406,
+      "eval_runtime": 3.7425,
+      "eval_samples_per_second": 180.626,
+      "eval_steps_per_second": 22.712,
       "step": 424
     },
     {
       "epoch": 3.0,
       "learning_rate": 9.850000000000001e-05,
-      "loss": 4.6838,
+      "loss": 4.6993,
       "step": 636
     },
     {
       "epoch": 3.0,
-      "eval_loss": 4.013933181762695,
-      "eval_runtime": 4.1662,
-      "eval_samples_per_second": 162.258,
-      "eval_steps_per_second": 20.402,
+      "eval_loss": 4.018372058868408,
+      "eval_runtime": 4.0424,
+      "eval_samples_per_second": 167.229,
+      "eval_steps_per_second": 21.027,
       "step": 636
     },
     {
       "epoch": 4.0,
       "learning_rate": 9.8e-05,
-      "loss": 3.9438,
+      "loss": 3.967,
       "step": 848
     },
     {
       "epoch": 4.0,
-      "eval_loss": 3.277362585067749,
-      "eval_runtime": 4.2258,
-      "eval_samples_per_second": 159.97,
-      "eval_steps_per_second": 20.115,
+      "eval_loss": 3.290557384490967,
+      "eval_runtime": 4.3933,
+      "eval_samples_per_second": 153.872,
+      "eval_steps_per_second": 19.348,
       "step": 848
     },
     {
       "epoch": 5.0,
       "learning_rate": 9.75e-05,
-      "loss": 3.2751,
+      "loss": 3.305,
       "step": 1060
     },
     {
       "epoch": 5.0,
-      "eval_loss": 2.62673282623291,
-      "eval_runtime": 4.5502,
-      "eval_samples_per_second": 148.565,
-      "eval_steps_per_second": 18.68,
+      "eval_loss": 2.6497254371643066,
+      "eval_runtime": 4.3562,
+      "eval_samples_per_second": 155.182,
+      "eval_steps_per_second": 19.513,
       "step": 1060
     },
     {
       "epoch": 6.0,
       "learning_rate": 9.7e-05,
-      "loss": 2.6863,
+      "loss": 2.7185,
       "step": 1272
     },
     {
       "epoch": 6.0,
-      "eval_loss": 2.0860772132873535,
-      "eval_runtime": 4.2538,
-      "eval_samples_per_second": 158.917,
-      "eval_steps_per_second": 19.982,
+      "eval_loss": 2.1158342361450195,
+      "eval_runtime": 4.5634,
+      "eval_samples_per_second": 148.135,
+      "eval_steps_per_second": 18.626,
       "step": 1272
     }
   ],
@@ -97,7 +97,7 @@
   "max_steps": 42400,
   "num_train_epochs": 200,
   "save_steps": 500,
-  "total_flos": 1855215309758544.0,
+  "total_flos": 1709449094176848.0,
   "trial_name": null,
   "trial_params": null
 }

checkpoint-1272/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b45ac36706a0f96e1fbfd751ddcc73ed3d4e95773bdb931420d6f73f3adff937
+oid sha256:4e7d47d9ef4da320ee46463f8716d365eb34b4bc5488090164af1bcf3ad60b35
 size 4155