Training in progress, step 1300, checkpoint

checkpoint-1300/README.md

@@ -0,0 +1,21 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: False
+- load_in_4bit: True
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: bfloat16
+### Framework versions
+
+
+- PEFT 0.6.0.dev0

checkpoint-1300/adapter_config.json

@@ -0,0 +1,21 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "codellama/CodeLlama-7b-Instruct-hf",
+  "bias": "lora_only",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 8,
+  "lora_dropout": 0.1,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

checkpoint-1300/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:810ba26da7359e3da161d3b327a7432cb722a3b5c108d9f68b1554a381be360e
+size 16822989

checkpoint-1300/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4f663172428509ffe6508c0873b117388ff1b2569ba0b2ecdfdac32713e62e2
+size 8555781

checkpoint-1300/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef5e2ef2e3e72a0a59c7bf4de078701fcdcc674a9e649ba2525b266ec0d1b780
+size 14575

checkpoint-1300/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91ee0df4f1fd05bf20d408222b4615835162bd845147eeca3ccae4196726b4d2
+size 627

checkpoint-1300/trainer_state.json

@@ -0,0 +1,285 @@
+{
+  "best_metric": 0.23951096832752228,
+  "best_model_checkpoint": "./text2sql/codellama_instruct_spider_e10/checkpoint-500",
+  "epoch": 4.803695150115473,
+  "eval_steps": 100,
+  "global_step": 1300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.08641975308642e-07,
+      "loss": 1.3897,
+      "step": 1
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.54320987654321e-05,
+      "loss": 1.3449,
+      "step": 50
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.08641975308642e-05,
+      "loss": 0.822,
+      "step": 100
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 0.5313276648521423,
+      "eval_runtime": 34.4984,
+      "eval_samples_per_second": 29.972,
+      "eval_steps_per_second": 3.768,
+      "step": 100
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.4342,
+      "step": 150
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.9251379038613084e-05,
+      "loss": 0.3014,
+      "step": 200
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 0.2763093411922455,
+      "eval_runtime": 34.3952,
+      "eval_samples_per_second": 30.062,
+      "eval_steps_per_second": 3.78,
+      "step": 200
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.826635145784082e-05,
+      "loss": 0.2401,
+      "step": 250
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.728132387706856e-05,
+      "loss": 0.2091,
+      "step": 300
+    },
+    {
+      "epoch": 1.11,
+      "eval_loss": 0.24685852229595184,
+      "eval_runtime": 34.5214,
+      "eval_samples_per_second": 29.952,
+      "eval_steps_per_second": 3.766,
+      "step": 300
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.1837,
+      "step": 350
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.5311268715524035e-05,
+      "loss": 0.1697,
+      "step": 400
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 0.2400880753993988,
+      "eval_runtime": 34.3334,
+      "eval_samples_per_second": 30.116,
+      "eval_steps_per_second": 3.786,
+      "step": 400
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.432624113475177e-05,
+      "loss": 0.1589,
+      "step": 450
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 4.334121355397952e-05,
+      "loss": 0.1495,
+      "step": 500
+    },
+    {
+      "epoch": 1.85,
+      "eval_loss": 0.23951096832752228,
+      "eval_runtime": 34.3515,
+      "eval_samples_per_second": 30.101,
+      "eval_steps_per_second": 3.784,
+      "step": 500
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.235618597320725e-05,
+      "loss": 0.1356,
+      "step": 550
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 4.1371158392434986e-05,
+      "loss": 0.1256,
+      "step": 600
+    },
+    {
+      "epoch": 2.22,
+      "eval_loss": 0.2525351345539093,
+      "eval_runtime": 34.4715,
+      "eval_samples_per_second": 29.996,
+      "eval_steps_per_second": 3.771,
+      "step": 600
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 4.0386130811662727e-05,
+      "loss": 0.1162,
+      "step": 650
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 3.940110323089047e-05,
+      "loss": 0.1097,
+      "step": 700
+    },
+    {
+      "epoch": 2.59,
+      "eval_loss": 0.2641294002532959,
+      "eval_runtime": 34.4325,
+      "eval_samples_per_second": 30.03,
+      "eval_steps_per_second": 3.776,
+      "step": 700
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.84160756501182e-05,
+      "loss": 0.1059,
+      "step": 750
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.743104806934594e-05,
+      "loss": 0.1107,
+      "step": 800
+    },
+    {
+      "epoch": 2.96,
+      "eval_loss": 0.2617126703262329,
+      "eval_runtime": 34.3286,
+      "eval_samples_per_second": 30.121,
+      "eval_steps_per_second": 3.787,
+      "step": 800
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 3.6446020488573684e-05,
+      "loss": 0.0929,
+      "step": 850
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.546099290780142e-05,
+      "loss": 0.0951,
+      "step": 900
+    },
+    {
+      "epoch": 3.33,
+      "eval_loss": 0.26825329661369324,
+      "eval_runtime": 34.3778,
+      "eval_samples_per_second": 30.078,
+      "eval_steps_per_second": 3.782,
+      "step": 900
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 3.447596532702916e-05,
+      "loss": 0.0885,
+      "step": 950
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.349093774625689e-05,
+      "loss": 0.0882,
+      "step": 1000
+    },
+    {
+      "epoch": 3.7,
+      "eval_loss": 0.28923752903938293,
+      "eval_runtime": 34.5013,
+      "eval_samples_per_second": 29.97,
+      "eval_steps_per_second": 3.768,
+      "step": 1000
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 3.2505910165484634e-05,
+      "loss": 0.0861,
+      "step": 1050
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.1520882584712375e-05,
+      "loss": 0.0818,
+      "step": 1100
+    },
+    {
+      "epoch": 4.06,
+      "eval_loss": 0.31338369846343994,
+      "eval_runtime": 34.3826,
+      "eval_samples_per_second": 30.073,
+      "eval_steps_per_second": 3.781,
+      "step": 1100
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 3.053585500394011e-05,
+      "loss": 0.0714,
+      "step": 1150
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 2.9550827423167847e-05,
+      "loss": 0.075,
+      "step": 1200
+    },
+    {
+      "epoch": 4.43,
+      "eval_loss": 0.2977830767631531,
+      "eval_runtime": 34.4574,
+      "eval_samples_per_second": 30.008,
+      "eval_steps_per_second": 3.773,
+      "step": 1200
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.8565799842395592e-05,
+      "loss": 0.0725,
+      "step": 1250
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 2.758077226162333e-05,
+      "loss": 0.0745,
+      "step": 1300
+    },
+    {
+      "epoch": 4.8,
+      "eval_loss": 0.3094833493232727,
+      "eval_runtime": 34.5265,
+      "eval_samples_per_second": 29.948,
+      "eval_steps_per_second": 3.765,
+      "step": 1300
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 2700,
+  "num_train_epochs": 10,
+  "save_steps": 100,
+  "total_flos": 3.968781131460772e+17,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-1300/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f894492930ff690bb0d1428d042fd66b325480fe96ad5ff777a305a69efec314
+size 4219