Training in progress, step 150, checkpoint

last-checkpoint/adapter_config.json

@@ -22,8 +22,8 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "k_proj"
+    "k_proj",
+    "q_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2c571c3d9bc97035512218123eb16c4a5ac39adeb400b8f87dca66f9daff0f4c
+oid sha256:abee9defa2f4343dad896a46c39f88dc5e2d2b650491c0c4e7d2178312355299
 size 527048968

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4bb2b5192e62889a659f5c32729030ddd962b6ab8d9ab4f6e8f06e1b2f7f86d0
+oid sha256:2e158809030e00ab57a06e7f58d9209551ec0bf02f8cf1e23626ea2cae56c82d
 size 1054135994

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:037d13720220086c05f76f1146cd4356e8b9d075b5d306338df00d366045e1c1
+oid sha256:5feb56512e955691dc9bb9a1e37b9dd590e06a961d7d94560b679e2730b03194
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6f4c00d522bdde510099aafe1617b13d114dce17a17b44e05876f016f4e4d7af
+oid sha256:8cf3f988e8fed2daa2e801eb1f19b681872781cf57f0fb7b896e859a12cfe2bb
 size 1064

last-checkpoint/trainer_state.json

@@ -1,302 +1,157 @@
 {
-  "best_metric": 0.7310147881507874,
-  "best_model_checkpoint": "./output/checkpoint-300",
-  "epoch": 13.043478260869565,
+  "best_metric": 0.7949807047843933,
+  "best_model_checkpoint": "./output/checkpoint-150",
+  "epoch": 6.521739130434782,
   "eval_steps": 150,
-  "global_step": 300,
+  "global_step": 150,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.43478260869565216,
-      "grad_norm": 1.5021581649780273,
+      "grad_norm": 1.5021635293960571,
       "learning_rate": 3e-06,
-      "loss": 0.9064,
+      "loss": 0.906,
       "step": 10
     },
     {
       "epoch": 0.8695652173913043,
-      "grad_norm": 1.6793084144592285,
+      "grad_norm": 1.687072992324829,
       "learning_rate": 6e-06,
-      "loss": 0.9027,
+      "loss": 0.9025,
       "step": 20
     },
     {
       "epoch": 1.3043478260869565,
-      "grad_norm": 1.7296501398086548,
+      "grad_norm": 1.7296812534332275,
       "learning_rate": 9e-06,
-      "loss": 0.9003,
+      "loss": 0.9001,
       "step": 30
     },
     {
       "epoch": 1.7391304347826086,
-      "grad_norm": 1.4458428621292114,
+      "grad_norm": 1.4459028244018555,
       "learning_rate": 1.2e-05,
-      "loss": 0.9092,
+      "loss": 0.909,
       "step": 40
     },
     {
       "epoch": 2.1739130434782608,
-      "grad_norm": 1.351685643196106,
+      "grad_norm": 1.351724624633789,
       "learning_rate": 1.5e-05,
-      "loss": 0.8363,
+      "loss": 0.8361,
       "step": 50
     },
     {
       "epoch": 2.608695652173913,
-      "grad_norm": 2.0312559604644775,
+      "grad_norm": 2.0466818809509277,
       "learning_rate": 1.8e-05,
       "loss": 0.8894,
       "step": 60
     },
     {
       "epoch": 3.0434782608695654,
-      "grad_norm": 1.4843778610229492,
+      "grad_norm": 1.4843230247497559,
       "learning_rate": 2.1e-05,
-      "loss": 0.8913,
+      "loss": 0.8912,
       "step": 70
     },
     {
       "epoch": 3.4782608695652173,
-      "grad_norm": 1.729393720626831,
+      "grad_norm": 1.7293753623962402,
       "learning_rate": 2.4e-05,
-      "loss": 0.8231,
+      "loss": 0.8237,
       "step": 80
     },
     {
       "epoch": 3.9130434782608696,
-      "grad_norm": 1.4239307641983032,
+      "grad_norm": 1.424095869064331,
       "learning_rate": 2.7000000000000002e-05,
-      "loss": 0.8528,
+      "loss": 0.8527,
       "step": 90
     },
     {
       "epoch": 4.3478260869565215,
-      "grad_norm": 1.3656258583068848,
+      "grad_norm": 1.3656634092330933,
       "learning_rate": 3e-05,
-      "loss": 0.865,
+      "loss": 0.8649,
       "step": 100
     },
     {
       "epoch": 4.782608695652174,
-      "grad_norm": 2.198690176010132,
+      "grad_norm": 2.198690891265869,
       "learning_rate": 2.999999702723963e-05,
-      "loss": 0.8228,
+      "loss": 0.8224,
       "step": 110
     },
     {
       "epoch": 5.217391304347826,
-      "grad_norm": 1.0728951692581177,
+      "grad_norm": 1.0726526975631714,
       "learning_rate": 2.9999988108959687e-05,
-      "loss": 0.7655,
+      "loss": 0.7651,
       "step": 120
     },
     {
       "epoch": 5.6521739130434785,
-      "grad_norm": 1.5677502155303955,
+      "grad_norm": 1.5673664808273315,
       "learning_rate": 2.9999973245163716e-05,
-      "loss": 0.7412,
+      "loss": 0.7415,
       "step": 130
     },
     {
       "epoch": 6.086956521739131,
-      "grad_norm": 1.9072229862213135,
+      "grad_norm": 1.9072725772857666,
       "learning_rate": 2.99999524358576e-05,
-      "loss": 0.7657,
+      "loss": 0.7655,
       "step": 140
     },
     {
       "epoch": 6.521739130434782,
-      "grad_norm": 1.1221002340316772,
+      "grad_norm": 1.1216552257537842,
       "learning_rate": 2.9999925681049593e-05,
       "loss": 0.7857,
       "step": 150
     },
     {
       "epoch": 6.521739130434782,
-      "eval_loss": 0.7962104678153992,
-      "eval_runtime": 0.4685,
-      "eval_samples_per_second": 21.344,
-      "eval_steps_per_second": 21.344,
+      "eval_loss": 0.7949807047843933,
+      "eval_runtime": 0.4775,
+      "eval_samples_per_second": 20.944,
+      "eval_steps_per_second": 20.944,
       "step": 150
     },
     {
       "epoch": 6.521739130434782,
-      "eval_loss": 0.8606523275375366,
-      "eval_runtime": 0.3993,
-      "eval_samples_per_second": 25.042,
-      "eval_steps_per_second": 25.042,
+      "eval_loss": 0.8609212040901184,
+      "eval_runtime": 0.417,
+      "eval_samples_per_second": 23.983,
+      "eval_steps_per_second": 23.983,
       "step": 150
     },
     {
       "epoch": 6.521739130434782,
-      "eval_loss": 0.7962104678153992,
-      "eval_runtime": 0.3965,
-      "eval_samples_per_second": 25.221,
-      "eval_steps_per_second": 25.221,
+      "eval_loss": 0.7949807047843933,
+      "eval_runtime": 0.4175,
+      "eval_samples_per_second": 23.952,
+      "eval_steps_per_second": 23.952,
       "step": 150
     },
     {
       "epoch": 6.521739130434782,
-      "eval_loss": 0.8606523275375366,
-      "eval_runtime": 0.4092,
-      "eval_samples_per_second": 24.439,
-      "eval_steps_per_second": 24.439,
+      "eval_loss": 0.8609212040901184,
+      "eval_runtime": 0.4278,
+      "eval_samples_per_second": 23.374,
+      "eval_steps_per_second": 23.374,
       "step": 150
     },
     {
       "epoch": 6.521739130434782,
-      "eval_loss": 0.8619200587272644,
-      "eval_runtime": 0.3955,
-      "eval_samples_per_second": 25.287,
-      "eval_steps_per_second": 25.287,
+      "eval_loss": 0.8608071208000183,
+      "eval_runtime": 0.4109,
+      "eval_samples_per_second": 24.338,
+      "eval_steps_per_second": 24.338,
       "step": 150
-    },
-    {
-      "epoch": 6.956521739130435,
-      "grad_norm": 1.540989637374878,
-      "learning_rate": 2.9999892980750297e-05,
-      "loss": 0.6586,
-      "step": 160
-    },
-    {
-      "epoch": 7.391304347826087,
-      "grad_norm": 1.345400094985962,
-      "learning_rate": 2.9999854334972675e-05,
-      "loss": 0.739,
-      "step": 170
-    },
-    {
-      "epoch": 7.826086956521739,
-      "grad_norm": 1.725967526435852,
-      "learning_rate": 2.999980974373204e-05,
-      "loss": 0.7291,
-      "step": 180
-    },
-    {
-      "epoch": 8.26086956521739,
-      "grad_norm": 1.5401579141616821,
-      "learning_rate": 2.9999759207046075e-05,
-      "loss": 0.6245,
-      "step": 190
-    },
-    {
-      "epoch": 8.695652173913043,
-      "grad_norm": 1.7425003051757812,
-      "learning_rate": 2.9999702724934804e-05,
-      "loss": 0.6765,
-      "step": 200
-    },
-    {
-      "epoch": 9.130434782608695,
-      "grad_norm": 1.0419254302978516,
-      "learning_rate": 2.999964029742062e-05,
-      "loss": 0.6524,
-      "step": 210
-    },
-    {
-      "epoch": 9.565217391304348,
-      "grad_norm": 1.2119251489639282,
-      "learning_rate": 2.9999571924528263e-05,
-      "loss": 0.5593,
-      "step": 220
-    },
-    {
-      "epoch": 10.0,
-      "grad_norm": 1.5276358127593994,
-      "learning_rate": 2.9999497606284837e-05,
-      "loss": 0.756,
-      "step": 230
-    },
-    {
-      "epoch": 10.434782608695652,
-      "grad_norm": 1.413993000984192,
-      "learning_rate": 2.9999417342719796e-05,
-      "loss": 0.7116,
-      "step": 240
-    },
-    {
-      "epoch": 10.869565217391305,
-      "grad_norm": 0.9750322699546814,
-      "learning_rate": 2.9999331133864956e-05,
-      "loss": 0.5897,
-      "step": 250
-    },
-    {
-      "epoch": 11.304347826086957,
-      "grad_norm": 1.1936322450637817,
-      "learning_rate": 2.9999238979754485e-05,
-      "loss": 0.6547,
-      "step": 260
-    },
-    {
-      "epoch": 11.73913043478261,
-      "grad_norm": 1.0489903688430786,
-      "learning_rate": 2.999914088042492e-05,
-      "loss": 0.6475,
-      "step": 270
-    },
-    {
-      "epoch": 12.173913043478262,
-      "grad_norm": 1.3138858079910278,
-      "learning_rate": 2.9999036835915132e-05,
-      "loss": 0.5937,
-      "step": 280
-    },
-    {
-      "epoch": 12.608695652173914,
-      "grad_norm": 1.0842424631118774,
-      "learning_rate": 2.9998926846266365e-05,
-      "loss": 0.6326,
-      "step": 290
-    },
-    {
-      "epoch": 13.043478260869565,
-      "grad_norm": 1.3883247375488281,
-      "learning_rate": 2.9998810911522213e-05,
-      "loss": 0.5806,
-      "step": 300
-    },
-    {
-      "epoch": 13.043478260869565,
-      "eval_loss": 0.7310147881507874,
-      "eval_runtime": 0.4443,
-      "eval_samples_per_second": 22.51,
-      "eval_steps_per_second": 22.51,
-      "step": 300
-    },
-    {
-      "epoch": 13.043478260869565,
-      "eval_loss": 0.8606523275375366,
-      "eval_runtime": 0.4404,
-      "eval_samples_per_second": 22.708,
-      "eval_steps_per_second": 22.708,
-      "step": 300
-    },
-    {
-      "epoch": 13.043478260869565,
-      "eval_loss": 0.7310147881507874,
-      "eval_runtime": 0.4391,
-      "eval_samples_per_second": 22.773,
-      "eval_steps_per_second": 22.773,
-      "step": 300
-    },
-    {
-      "epoch": 13.043478260869565,
-      "eval_loss": 0.7758375406265259,
-      "eval_runtime": 0.4396,
-      "eval_samples_per_second": 22.75,
-      "eval_steps_per_second": 22.75,
-      "step": 300
-    },
-    {
-      "epoch": 13.043478260869565,
-      "eval_loss": 0.860957145690918,
-      "eval_runtime": 0.4444,
-      "eval_samples_per_second": 22.5,
-      "eval_steps_per_second": 22.5,
-      "step": 300
     }
   ],
   "logging_steps": 10,
@@ -316,7 +171,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 7724643094462464.0,
+  "total_flos": 3894839614291968.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:003a11451b42f43906137235259ca60836693a9072ec318195749b927967a9e1
+oid sha256:1943b73dd12e0d46a3b26b62916dc0f34ed8d7e7a1a2985c54b619124519181f
 size 5368