SARG-ai/test-atomizer-finetune

adapter_config.json

@@ -23,13 +23,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
+    "k_proj",
+    "gate_proj",
     "v_proj",
-    "o_proj",
-    "up_proj",
+    "q_proj",
     "down_proj",
-    "k_proj",
-    "gate_proj"
+    "up_proj",
+    "o_proj"
   ],
   "task_type": null,
   "use_dora": false,

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 1.0,
     "total_flos": 0.0,
-    "train_loss": 1.770367290686678,
-    "train_runtime": 102.5984,
-    "train_samples_per_second": 1.579,
-    "train_steps_per_second": 0.789
+    "train_loss": 0.857978654280305,
+    "train_runtime": 207.5744,
+    "train_samples_per_second": 1.542,
+    "train_steps_per_second": 0.771
 }

model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9c994a40e131080d93cbef61e284b793761b451006a34720672440c80096ccc7
+oid sha256:1f6f4f42979e9f3f64e43d9c75a03812ee24c43bc8c26db10cde02404e09e2bf
 size 4877660776

model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c7a96313da3d2879efc2fe61ac39e0cbe81b54452ab8013ae6746236840a7fb0
+oid sha256:f6f6f7301d83ebd777d4cfdcf31ae9ee8427fa8a34e4ed676a762d87ef75d528
 size 4932751008

model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:43bfd5cad7d2fabaa3e01cc208c1e888fe999b67c9fa4a2c1ed8c57eedfabc0c
+oid sha256:16754098bd73a0a80e657c0190944bcf7ec5ec18cf76106462e8fb195ab53577
 size 4330865200

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 1.0,
     "total_flos": 0.0,
-    "train_loss": 1.770367290686678,
-    "train_runtime": 102.5984,
-    "train_samples_per_second": 1.579,
-    "train_steps_per_second": 0.789
+    "train_loss": 0.857978654280305,
+    "train_runtime": 207.5744,
+    "train_samples_per_second": 1.542,
+    "train_steps_per_second": 0.771
 }

trainer_state.json

@@ -3,86 +3,149 @@
   "best_model_checkpoint": null,
   "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 81,
+  "global_step": 160,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.12345679012345678,
-      "grad_norm": 12.38063907623291,
-      "learning_rate": 8.765432098765433e-05,
-      "loss": 7.922,
+      "epoch": 0.0625,
+      "grad_norm": 15.055679321289062,
+      "learning_rate": 9.375e-05,
+      "loss": 8.1614,
       "step": 10
     },
     {
-      "epoch": 0.24691358024691357,
-      "grad_norm": 13.536417007446289,
-      "learning_rate": 7.530864197530865e-05,
-      "loss": 4.6077,
+      "epoch": 0.125,
+      "grad_norm": 14.071452140808105,
+      "learning_rate": 8.75e-05,
+      "loss": 4.3181,
       "step": 20
     },
     {
-      "epoch": 0.37037037037037035,
-      "grad_norm": 2.8598146438598633,
-      "learning_rate": 6.296296296296296e-05,
-      "loss": 1.327,
+      "epoch": 0.1875,
+      "grad_norm": 0.31610435247421265,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 0.6166,
       "step": 30
     },
     {
-      "epoch": 0.49382716049382713,
-      "grad_norm": 0.32510578632354736,
-      "learning_rate": 5.061728395061729e-05,
-      "loss": 0.0959,
+      "epoch": 0.25,
+      "grad_norm": 0.2891228199005127,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 0.0491,
       "step": 40
     },
     {
-      "epoch": 0.6172839506172839,
-      "grad_norm": 0.2507149875164032,
-      "learning_rate": 3.82716049382716e-05,
-      "loss": 0.1085,
+      "epoch": 0.3125,
+      "grad_norm": 0.4224626421928406,
+      "learning_rate": 6.875e-05,
+      "loss": 0.0563,
       "step": 50
     },
     {
-      "epoch": 0.7407407407407407,
-      "grad_norm": 0.2675971984863281,
-      "learning_rate": 2.5925925925925925e-05,
-      "loss": 0.0834,
+      "epoch": 0.375,
+      "grad_norm": 0.2085612267255783,
+      "learning_rate": 6.25e-05,
+      "loss": 0.0456,
       "step": 60
     },
     {
-      "epoch": 0.8641975308641975,
-      "grad_norm": 0.22273920476436615,
-      "learning_rate": 1.3580246913580247e-05,
-      "loss": 0.1037,
+      "epoch": 0.4375,
+      "grad_norm": 0.19864903390407562,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 0.0518,
       "step": 70
     },
     {
-      "epoch": 0.9876543209876543,
-      "grad_norm": 0.20021985471248627,
-      "learning_rate": 1.234567901234568e-06,
-      "loss": 0.0836,
+      "epoch": 0.5,
+      "grad_norm": 0.1632135957479477,
+      "learning_rate": 5e-05,
+      "loss": 0.051,
       "step": 80
     },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.22292855381965637,
+      "learning_rate": 4.375e-05,
+      "loss": 0.0482,
+      "step": 90
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.2120644748210907,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.0488,
+      "step": 100
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.20597754418849945,
+      "learning_rate": 3.125e-05,
+      "loss": 0.0484,
+      "step": 110
+    },
+    {
+      "epoch": 0.6875,
+      "eval_runtime": 16.2029,
+      "eval_samples_per_second": 4.937,
+      "eval_steps_per_second": 0.309,
+      "step": 110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.25182241201400757,
+      "learning_rate": 2.5e-05,
+      "loss": 0.0453,
+      "step": 120
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 0.1939142495393753,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.0463,
+      "step": 130
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.1711636185646057,
+      "learning_rate": 1.25e-05,
+      "loss": 0.0528,
+      "step": 140
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.17180846631526947,
+      "learning_rate": 6.25e-06,
+      "loss": 0.0453,
+      "step": 150
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.17636561393737793,
+      "learning_rate": 0.0,
+      "loss": 0.0427,
+      "step": 160
+    },
     {
       "epoch": 1.0,
-      "eval_runtime": 8.9544,
-      "eval_samples_per_second": 4.579,
-      "eval_steps_per_second": 0.335,
-      "step": 81
+      "eval_runtime": 14.9912,
+      "eval_samples_per_second": 5.336,
+      "eval_steps_per_second": 0.334,
+      "step": 160
     },
     {
       "epoch": 1.0,
-      "step": 81,
+      "step": 160,
       "total_flos": 0.0,
-      "train_loss": 1.770367290686678,
-      "train_runtime": 102.5984,
-      "train_samples_per_second": 1.579,
-      "train_steps_per_second": 0.789
+      "train_loss": 0.857978654280305,
+      "train_runtime": 207.5744,
+      "train_samples_per_second": 1.542,
+      "train_steps_per_second": 0.771
     }
   ],
   "logging_steps": 10,
-  "max_steps": 81,
+  "max_steps": 160,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 500,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:551092e3a36216a13e25d403df5e47bdf244b4e4c6981cf6cd7add30f473525c
+oid sha256:47242d08a39472cb54f012c57cadf26fb6c49f357a307054e2c99ed22dbc1521
 size 5432