Training in progress, step 30000

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bfba69a38235688144762f0028225bf75ceb8ed54f934556abed7f29749c0a1c
+oid sha256:9537b3a7e6055deff50c06679b380e3a915c90d647696e0b31fc0265c367cbbc
 size 500897101

{checkpoint-26000 → checkpoint-29000/adapter_model}/adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:db604a690ace9df3e57323e7728a62d95ca6fe945a7a4f0390b60803c2b0d0f0
+oid sha256:bfba69a38235688144762f0028225bf75ceb8ed54f934556abed7f29749c0a1c
 size 500897101

{checkpoint-26000/adapter_model → checkpoint-30000}/adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:db604a690ace9df3e57323e7728a62d95ca6fe945a7a4f0390b60803c2b0d0f0
+oid sha256:9537b3a7e6055deff50c06679b380e3a915c90d647696e0b31fc0265c367cbbc
 size 500897101

{checkpoint-26000 → checkpoint-30000}/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:27e0a8d1f02955048c81208b54b8e952c63d97fda72aa75ae96fcfc8f35af7cb
+oid sha256:be1c4c976ca70fca137b1074a48111db76f8417f793dd8b169c29e76d799f173
 size 1001723453

{checkpoint-26000 → checkpoint-30000}/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:23c71a7c504ad8a21f65a57923dde0238456dda222b9af4b24d2bfc5f74e422c
+oid sha256:e5aed0374c0ea79a6b6022726183fe49008353437a1d89b31492e7e657ff784f
 size 14575

{checkpoint-26000 → checkpoint-30000}/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8a9f0595f0fbccf1fcdc4c120788726227aa9e6442e4233fc3c71485b74b1e4e
+oid sha256:3860c8699ca30c96bb0a9da34e9e209f4fa8a3c56eb8f9c2ce426750974e9aa6
 size 627

{checkpoint-26000 → checkpoint-30000}/trainer_state.json

@@ -1,8 +1,8 @@
 {
-  "best_metric": 0.6026987433433533,
-  "best_model_checkpoint": "./qlora-out/checkpoint-26000",
-  "epoch": 0.9693896573580404,
-  "global_step": 26000,
+  "best_metric": 0.5869857668876648,
+  "best_model_checkpoint": "./qlora-out/checkpoint-30000",
+  "epoch": 1.1185265277208158,
+  "global_step": 30000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -1774,11 +1774,283 @@
       "eval_samples_per_second": 0.419,
       "eval_steps_per_second": 0.419,
       "step": 26000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00015244025404312974,
+      "loss": 0.6015,
+      "step": 26100
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015210736612567588,
+      "loss": 0.5914,
+      "step": 26200
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015177368367176616,
+      "loss": 0.5799,
+      "step": 26300
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015143921176941205,
+      "loss": 0.6037,
+      "step": 26400
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00015110395551866255,
+      "loss": 0.5876,
+      "step": 26500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001507679200315264,
+      "loss": 0.5973,
+      "step": 26600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00015043111043189423,
+      "loss": 0.5957,
+      "step": 26700
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00015009353185546046,
+      "loss": 0.5696,
+      "step": 26800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00014975518944964478,
+      "loss": 0.5523,
+      "step": 26900
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0001494160883735139,
+      "loss": 0.5144,
+      "step": 27000
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 0.5985096096992493,
+      "eval_runtime": 1314.8131,
+      "eval_samples_per_second": 0.412,
+      "eval_steps_per_second": 0.412,
+      "step": 27000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00014907623379770263,
+      "loss": 0.5743,
+      "step": 27100
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00014873563090433547,
+      "loss": 0.5095,
+      "step": 27200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00014839428488694706,
+      "loss": 0.5391,
+      "step": 27300
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00014805220095040334,
+      "loss": 0.5532,
+      "step": 27400
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014770938431082212,
+      "loss": 0.536,
+      "step": 27500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014736584019549342,
+      "loss": 0.5204,
+      "step": 27600
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014702157384279997,
+      "loss": 0.5026,
+      "step": 27700
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001466765905021371,
+      "loss": 0.5319,
+      "step": 27800
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00014633089543383295,
+      "loss": 0.5112,
+      "step": 27900
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00014598449390906804,
+      "loss": 0.5146,
+      "step": 28000
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 0.5959522128105164,
+      "eval_runtime": 1288.6066,
+      "eval_samples_per_second": 0.421,
+      "eval_steps_per_second": 0.421,
+      "step": 28000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00014563739120979497,
+      "loss": 0.5262,
+      "step": 28100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00014528959262865798,
+      "loss": 0.5082,
+      "step": 28200
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00014494110346891206,
+      "loss": 0.5094,
+      "step": 28300
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00014459192904434226,
+      "loss": 0.5012,
+      "step": 28400
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0001442420746791826,
+      "loss": 0.4946,
+      "step": 28500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00014389154570803477,
+      "loss": 0.5138,
+      "step": 28600
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.000143540347475787,
+      "loss": 0.5082,
+      "step": 28700
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001431884853375325,
+      "loss": 0.4842,
+      "step": 28800
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0001428359646584876,
+      "loss": 0.5143,
+      "step": 28900
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014248279081391022,
+      "loss": 0.5029,
+      "step": 29000
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 0.5910914540290833,
+      "eval_runtime": 1278.8257,
+      "eval_samples_per_second": 0.424,
+      "eval_steps_per_second": 0.424,
+      "step": 29000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014212896918901774,
+      "loss": 0.5003,
+      "step": 29100
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00014177450517890503,
+      "loss": 0.5102,
+      "step": 29200
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0001414194041884619,
+      "loss": 0.524,
+      "step": 29300
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001410636716322911,
+      "loss": 0.5168,
+      "step": 29400
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001407073129346254,
+      "loss": 0.514,
+      "step": 29500
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00014035033352924502,
+      "loss": 0.5084,
+      "step": 29600
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0001399927388593948,
+      "loss": 0.5203,
+      "step": 29700
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00013963453437770119,
+      "loss": 0.5226,
+      "step": 29800
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.000139275725546089,
+      "loss": 0.5055,
+      "step": 29900
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00013891631783569838,
+      "loss": 0.5303,
+      "step": 30000
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 0.5869857668876648,
+      "eval_runtime": 1272.0282,
+      "eval_samples_per_second": 0.426,
+      "eval_steps_per_second": 0.426,
+      "step": 30000
     }
   ],
   "max_steps": 80463,
   "num_train_epochs": 3,
-  "total_flos": 7.289703933515858e+18,
+  "total_flos": 8.413138321233592e+18,
   "trial_name": null,
   "trial_params": null
 }