Upload folder using huggingface_hub

checkpoint-step-500/adapter_config.json

@@ -34,13 +34,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
+    "gate_proj",
+    "k_proj",
     "v_proj",
     "down_proj",
     "o_proj",
-    "q_proj",
     "up_proj",
-    "gate_proj",
-    "k_proj"
+    "q_proj"
   ],
   "target_parameters": null,
   "task_type": "CAUSAL_LM",

checkpoint-step-500/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c863fc283ce45c9f631a699242fa10aedebe6e12b2d5b36c33198de2e5bd85a4
+oid sha256:a387be8c8d7260b40b7435544ff33549a9b554d949662c2dcd4a6e9454c3ac1a
 size 239536272

checkpoint-step-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:518a6b042e03bc5532d1f29e406467f275e01ce34e285daa710f09ee75d71e21
+oid sha256:62efc2e5e7c3fde724f3e0082f7b5bed74ea95019d8c6f3e9af7cfb5cd119967
 size 122203333

checkpoint-step-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d1459e426da085c489658844b6eaef146617542828cbc7b74d93aa3bb305e910
+oid sha256:690ef27a29abb181e57e795f5ffd494219b89b6cc90e622e8445b188e53b77d6
 size 1465

checkpoint-step-500/tokenizer_config.json

@@ -5,7 +5,7 @@
   "clean_up_tokenization_spaces": false,
   "eos_token": "<|endoftext|>",
   "errors": "replace",
-  "is_local": false,
+  "is_local": true,
   "model_max_length": 32768,
   "pad_token": "<|PAD_TOKEN|>",
   "padding_side": "right",

checkpoint-step-500/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_global_step": 500,
-  "best_metric": 1.0487685203552246,
-  "best_model_checkpoint": "./outputs/checkpoint-500",
+  "best_global_step": 400,
+  "best_metric": 0.9676186442375183,
+  "best_model_checkpoint": "./outputs/checkpoint-400",
   "epoch": 0.9398496240601504,
   "eval_steps": 100,
   "global_step": 500,
@@ -11,119 +11,119 @@
   "log_history": [
     {
       "epoch": 0.09398496240601503,
-      "grad_norm": 0.49556756019592285,
+      "grad_norm": 0.9101622104644775,
       "learning_rate": 9.8e-05,
-      "loss": 1.8540109252929688,
+      "loss": 0.7848518371582032,
       "step": 50
     },
     {
       "epoch": 0.18796992481203006,
-      "grad_norm": 0.4922962486743927,
-      "learning_rate": 9.907020872865275e-05,
-      "loss": 1.2944253540039063,
+      "grad_norm": 0.8308438062667847,
+      "learning_rate": 9.683053040103493e-05,
+      "loss": 0.7757094573974609,
       "step": 100
     },
     {
       "epoch": 0.18796992481203006,
-      "eval_loss": 1.216850757598877,
-      "eval_runtime": 83.4733,
-      "eval_samples_per_second": 5.666,
-      "eval_steps_per_second": 1.426,
+      "eval_loss": 0.9840108752250671,
+      "eval_runtime": 81.3507,
+      "eval_samples_per_second": 5.814,
+      "eval_steps_per_second": 1.463,
       "step": 100
     },
     {
       "epoch": 0.2819548872180451,
-      "grad_norm": 0.5891006588935852,
-      "learning_rate": 9.81214421252372e-05,
-      "loss": 1.2092308807373047,
+      "grad_norm": 1.0474584102630615,
+      "learning_rate": 9.359637774902976e-05,
+      "loss": 0.7674188995361328,
       "step": 150
     },
     {
       "epoch": 0.37593984962406013,
-      "grad_norm": 0.5563158392906189,
-      "learning_rate": 9.717267552182164e-05,
-      "loss": 1.1433011627197265,
+      "grad_norm": 0.8780233860015869,
+      "learning_rate": 9.036222509702458e-05,
+      "loss": 0.7395731353759766,
       "step": 200
     },
     {
       "epoch": 0.37593984962406013,
-      "eval_loss": 1.1367270946502686,
-      "eval_runtime": 82.9022,
-      "eval_samples_per_second": 5.706,
-      "eval_steps_per_second": 1.435,
+      "eval_loss": 0.9940316081047058,
+      "eval_runtime": 80.2928,
+      "eval_samples_per_second": 5.891,
+      "eval_steps_per_second": 1.482,
       "step": 200
     },
     {
       "epoch": 0.4699248120300752,
-      "grad_norm": 0.5556582808494568,
-      "learning_rate": 9.622390891840608e-05,
-      "loss": 1.1522552490234375,
+      "grad_norm": 0.9176077246665955,
+      "learning_rate": 8.712807244501941e-05,
+      "loss": 0.7627955627441406,
       "step": 250
     },
     {
       "epoch": 0.5639097744360902,
-      "grad_norm": 0.5758287906646729,
-      "learning_rate": 9.527514231499052e-05,
-      "loss": 1.0814830780029296,
+      "grad_norm": 0.8625491857528687,
+      "learning_rate": 8.389391979301423e-05,
+      "loss": 0.7164928436279296,
       "step": 300
     },
     {
       "epoch": 0.5639097744360902,
-      "eval_loss": 1.1028286218643188,
-      "eval_runtime": 82.616,
-      "eval_samples_per_second": 5.725,
-      "eval_steps_per_second": 1.44,
+      "eval_loss": 0.9823606014251709,
+      "eval_runtime": 80.8571,
+      "eval_samples_per_second": 5.85,
+      "eval_steps_per_second": 1.472,
       "step": 300
     },
     {
       "epoch": 0.6578947368421053,
-      "grad_norm": 0.5769486427307129,
-      "learning_rate": 9.432637571157496e-05,
-      "loss": 1.124346466064453,
+      "grad_norm": 0.8384292125701904,
+      "learning_rate": 8.065976714100906e-05,
+      "loss": 0.73987548828125,
       "step": 350
     },
     {
       "epoch": 0.7518796992481203,
-      "grad_norm": 0.5422045588493347,
-      "learning_rate": 9.337760910815939e-05,
-      "loss": 1.0794380950927733,
+      "grad_norm": 0.8384448885917664,
+      "learning_rate": 7.742561448900388e-05,
+      "loss": 0.7277918243408203,
       "step": 400
     },
     {
       "epoch": 0.7518796992481203,
-      "eval_loss": 1.072836995124817,
-      "eval_runtime": 82.9563,
-      "eval_samples_per_second": 5.702,
-      "eval_steps_per_second": 1.434,
+      "eval_loss": 0.9676186442375183,
+      "eval_runtime": 80.4443,
+      "eval_samples_per_second": 5.88,
+      "eval_steps_per_second": 1.479,
       "step": 400
     },
     {
       "epoch": 0.8458646616541353,
-      "grad_norm": 0.5861876606941223,
-      "learning_rate": 9.242884250474384e-05,
-      "loss": 1.0750784301757812,
+      "grad_norm": 0.866162121295929,
+      "learning_rate": 7.41914618369987e-05,
+      "loss": 0.7323916625976562,
       "step": 450
     },
     {
       "epoch": 0.9398496240601504,
-      "grad_norm": 0.5859329700469971,
-      "learning_rate": 9.148007590132827e-05,
-      "loss": 1.083390655517578,
+      "grad_norm": 0.8226180076599121,
+      "learning_rate": 7.095730918499353e-05,
+      "loss": 0.7260687255859375,
       "step": 500
     },
     {
       "epoch": 0.9398496240601504,
-      "eval_loss": 1.0487685203552246,
-      "eval_runtime": 82.9841,
-      "eval_samples_per_second": 5.7,
-      "eval_steps_per_second": 1.434,
+      "eval_loss": 0.9678767919540405,
+      "eval_runtime": 80.5208,
+      "eval_samples_per_second": 5.874,
+      "eval_steps_per_second": 1.478,
       "step": 500
     }
   ],
   "logging_steps": 50,
-  "max_steps": 5320,
+  "max_steps": 1596,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 3,
   "save_steps": 100,
   "stateful_callbacks": {
     "TrainerControl": {

checkpoint-step-500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bc05806d2c86681ef6c541c75942126f28b3a31aec0d903899ed9bf1ea303bac
+oid sha256:3c88f5ec73ab38f2f0a5bb321f3c58aaa6f716ae0e803b69ff493f9e0c43f221
 size 5713