Upload folder using huggingface_hub

adapter_config.json

@@ -25,13 +25,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
+    "gate_proj",
+    "k_proj",
     "v_proj",
-    "q_proj",
     "down_proj",
+    "q_proj",
     "up_proj",
-    "gate_proj",
-    "o_proj",
-    "k_proj"
+    "o_proj"
   ],
   "target_parameters": null,
   "task_type": "CAUSAL_LM",

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cce46b99801974dee4d50e7762369c85db39e3719f7e13f39c7757ed0201a65c
+oid sha256:a4fd2f2f24b663f9f3697c74aa5c31dd0b09aa20161cfd75d3472b50e1bbf472
 size 664584480

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a59b6490efd84469e8acf94b22f2093427cf545527a325d87b0512d1162e4bb2
+oid sha256:7cfd1aad66e313c4b2faa3c98a87e667e8808b634d9d98b7f19bc2a73239fdec
 size 1329377575

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:447f6d9c3def923b2023bfae8d2c470e245de58e058e98ae4722cc77fe074f8b
+oid sha256:012319d9d7b07efb800bfdc5b30f3b33091204a1f615665fe2368e0bd6978503
 size 14645

trainer_state.json

@@ -1,6 +1,6 @@
 {
   "best_global_step": 100,
-  "best_metric": 0.09546195715665817,
+  "best_metric": 0.09543681889772415,
   "best_model_checkpoint": "/content/models/gemma_qlora_lmh/checkpoint-100",
   "epoch": 1.7008547008547008,
   "eval_steps": 20,
@@ -10,108 +10,108 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "entropy": 2.4611108183860777,
+      "entropy": 2.4621588349342347,
       "epoch": 0.3418803418803419,
-      "grad_norm": 6.039472579956055,
+      "grad_norm": 6.012325763702393,
       "learning_rate": 8.389830508474577e-06,
-      "loss": 0.3829,
-      "mean_token_accuracy": 0.8768418416380882,
+      "loss": 0.3823,
+      "mean_token_accuracy": 0.8764635115861893,
       "num_tokens": 113164.0,
       "step": 20
     },
     {
       "epoch": 0.3418803418803419,
-      "eval_entropy": 2.3045855587364263,
-      "eval_loss": 0.14144526422023773,
-      "eval_mean_token_accuracy": 0.9522208352374215,
+      "eval_entropy": 2.307236980169247,
+      "eval_loss": 0.1412736475467682,
+      "eval_mean_token_accuracy": 0.9523680826537629,
       "eval_num_tokens": 113164.0,
-      "eval_runtime": 46.5777,
-      "eval_samples_per_second": 39.933,
-      "eval_steps_per_second": 2.512,
+      "eval_runtime": 46.9847,
+      "eval_samples_per_second": 39.587,
+      "eval_steps_per_second": 2.49,
       "step": 20
     },
     {
-      "entropy": 2.3063884317874908,
+      "entropy": 2.30367816388607,
       "epoch": 0.6837606837606838,
-      "grad_norm": 2.049039840698242,
+      "grad_norm": 2.036860466003418,
       "learning_rate": 6.694915254237288e-06,
-      "loss": 0.1361,
-      "mean_token_accuracy": 0.9577587321400642,
+      "loss": 0.1364,
+      "mean_token_accuracy": 0.9568088531494141,
       "num_tokens": 225335.0,
       "step": 40
     },
     {
       "epoch": 0.6837606837606838,
-      "eval_entropy": 2.2767278709982195,
-      "eval_loss": 0.11506718397140503,
-      "eval_mean_token_accuracy": 0.9626757140852448,
+      "eval_entropy": 2.2767869560127583,
+      "eval_loss": 0.11478288471698761,
+      "eval_mean_token_accuracy": 0.9625477276296697,
       "eval_num_tokens": 225335.0,
-      "eval_runtime": 45.3833,
-      "eval_samples_per_second": 40.984,
-      "eval_steps_per_second": 2.578,
+      "eval_runtime": 45.5897,
+      "eval_samples_per_second": 40.799,
+      "eval_steps_per_second": 2.566,
       "step": 40
     },
     {
-      "entropy": 2.299742362438104,
+      "entropy": 2.2971509970151462,
       "epoch": 1.017094017094017,
-      "grad_norm": 2.2758777141571045,
+      "grad_norm": 2.243170976638794,
       "learning_rate": 5e-06,
-      "loss": 0.1132,
-      "mean_token_accuracy": 0.9659271622315432,
+      "loss": 0.1134,
+      "mean_token_accuracy": 0.966560884928092,
       "num_tokens": 330390.0,
       "step": 60
     },
     {
       "epoch": 1.017094017094017,
-      "eval_entropy": 2.2947008232785087,
-      "eval_loss": 0.10887181758880615,
-      "eval_mean_token_accuracy": 0.9646475982462239,
+      "eval_entropy": 2.289694781996246,
+      "eval_loss": 0.10902266204357147,
+      "eval_mean_token_accuracy": 0.9641746343710483,
       "eval_num_tokens": 330390.0,
-      "eval_runtime": 45.9184,
-      "eval_samples_per_second": 40.507,
-      "eval_steps_per_second": 2.548,
+      "eval_runtime": 46.8001,
+      "eval_samples_per_second": 39.744,
+      "eval_steps_per_second": 2.5,
       "step": 60
     },
     {
-      "entropy": 2.277792435884476,
+      "entropy": 2.272606986761093,
       "epoch": 1.358974358974359,
-      "grad_norm": 2.237907886505127,
+      "grad_norm": 2.2923057079315186,
       "learning_rate": 3.305084745762712e-06,
-      "loss": 0.0848,
-      "mean_token_accuracy": 0.9723074913024903,
+      "loss": 0.0845,
+      "mean_token_accuracy": 0.9724922418594361,
       "num_tokens": 440357.0,
       "step": 80
     },
     {
       "epoch": 1.358974358974359,
-      "eval_entropy": 2.2612023394331975,
-      "eval_loss": 0.10450778901576996,
-      "eval_mean_token_accuracy": 0.9650997463454548,
+      "eval_entropy": 2.2530585782140746,
+      "eval_loss": 0.1047038808465004,
+      "eval_mean_token_accuracy": 0.9652910543303205,
       "eval_num_tokens": 440357.0,
-      "eval_runtime": 46.2067,
-      "eval_samples_per_second": 40.254,
-      "eval_steps_per_second": 2.532,
+      "eval_runtime": 46.2572,
+      "eval_samples_per_second": 40.21,
+      "eval_steps_per_second": 2.529,
       "step": 80
     },
     {
-      "entropy": 2.2707396924495695,
+      "entropy": 2.26450654566288,
       "epoch": 1.7008547008547008,
-      "grad_norm": 1.7053431272506714,
+      "grad_norm": 1.6888355016708374,
       "learning_rate": 1.6101694915254237e-06,
       "loss": 0.0715,
-      "mean_token_accuracy": 0.9734432741999626,
+      "mean_token_accuracy": 0.9734723582863808,
       "num_tokens": 552807.0,
       "step": 100
     },
     {
       "epoch": 1.7008547008547008,
-      "eval_entropy": 2.2446225402701616,
-      "eval_loss": 0.09546195715665817,
-      "eval_mean_token_accuracy": 0.9683777588045496,
+      "eval_entropy": 2.240677540118878,
+      "eval_loss": 0.09543681889772415,
+      "eval_mean_token_accuracy": 0.9683268676456224,
       "eval_num_tokens": 552807.0,
-      "eval_runtime": 45.8097,
-      "eval_samples_per_second": 40.603,
-      "eval_steps_per_second": 2.554,
+      "eval_runtime": 46.1331,
+      "eval_samples_per_second": 40.318,
+      "eval_steps_per_second": 2.536,
       "step": 100
     }
   ],

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:062d76cca895bbae1ddc531b525e8d06e13c9a529c7c97a8a71d7d66d737ee28
-size 6225
+oid sha256:f9399b718a9dab6f82ca03abb475407342319536237c41ccfc6081473e94f69b
+size 6289