2023-08-03 19:29:40 Autosave for checkpoint additions

{checkpoint-7000 → checkpoint-10000/adapter_model}/adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:157f6b6252bea0a483afac1b940a1e1d1fef2b0645bc8ee85e303857453144a6
+oid sha256:2de4db5b573eb8f174ab1eddb1a6ebefdbbe9493c4faf404321a5e114db196a0
 size 500897101

checkpoint-7000/adapter_model/README.md

@@ -1,20 +0,0 @@
----
-library_name: peft
----
-## Training procedure
-
-
-The following `bitsandbytes` quantization config was used during training:
-- 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 0.5.0.dev0

checkpoint-7000/adapter_model/adapter_config.json

@@ -1,26 +0,0 @@
-{
-  "auto_mapping": null,
-  "base_model_name_or_path": "/workspace/webui/models/TheBloke_Llama-2-13B-fp16",
-  "bias": "none",
-  "fan_in_fan_out": null,
-  "inference_mode": true,
-  "init_lora_weights": true,
-  "layers_pattern": null,
-  "layers_to_transform": null,
-  "lora_alpha": 16,
-  "lora_dropout": 0.05,
-  "modules_to_save": null,
-  "peft_type": "LORA",
-  "r": 32,
-  "revision": null,
-  "target_modules": [
-    "v_proj",
-    "k_proj",
-    "up_proj",
-    "down_proj",
-    "gate_proj",
-    "o_proj",
-    "q_proj"
-  ],
-  "task_type": "CAUSAL_LM"
-}

checkpoint-7000/adapter_model/adapter_model.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:157f6b6252bea0a483afac1b940a1e1d1fef2b0645bc8ee85e303857453144a6
-size 500897101

checkpoint-7000/optimizer.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2c933df6e99a74286946dcce224ad5e51568c15422a85cf0a7860884858dac50
-size 1001723453

checkpoint-7000/rng_state.pth

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b5098d77b456bcfbf087e59af41ffe1580ac2d35a275a4e7d1f88650e660b908
-size 14575

checkpoint-7000/scheduler.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5bf2d5f7ab561d255de1fb492947daa4108b2df0f44366f3347cba87aa6df44b
-size 627

checkpoint-7000/trainer_state.json

@@ -1,492 +0,0 @@
-{
-  "best_metric": 0.703677773475647,
-  "best_model_checkpoint": "./qlora-out/checkpoint-7000",
-  "epoch": 0.260989523134857,
-  "global_step": 7000,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.0,
-      "learning_rate": 0.00019999938245325715,
-      "loss": 0.9023,
-      "step": 100
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00019999724773356797,
-      "loss": 0.8027,
-      "step": 200
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.0001999935882494411,
-      "loss": 0.8041,
-      "step": 300
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 0.00019998840405667672,
-      "loss": 0.7944,
-      "step": 400
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.00019998169523432365,
-      "loss": 0.81,
-      "step": 500
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 0.0001999734618846785,
-      "loss": 0.7855,
-      "step": 600
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00019996370413328385,
-      "loss": 0.7849,
-      "step": 700
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00019995242212892653,
-      "loss": 0.7564,
-      "step": 800
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 0.00019993961604363532,
-      "loss": 0.7724,
-      "step": 900
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00019992528607267815,
-      "loss": 0.7308,
-      "step": 1000
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 0.7677998542785645,
-      "eval_runtime": 1774.3517,
-      "eval_samples_per_second": 0.305,
-      "eval_steps_per_second": 0.305,
-      "step": 1000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.0001999094324345594,
-      "loss": 0.7844,
-      "step": 1100
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 0.00019989205537101633,
-      "loss": 0.7668,
-      "step": 1200
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00019987315514701553,
-      "loss": 0.7727,
-      "step": 1300
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 0.00019985273205074878,
-      "loss": 0.7467,
-      "step": 1400
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00019983078639362883,
-      "loss": 0.7516,
-      "step": 1500
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.00019980731851028445,
-      "loss": 0.7267,
-      "step": 1600
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 0.0001997823287585554,
-      "loss": 0.7632,
-      "step": 1700
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.000199755817519487,
-      "loss": 0.7392,
-      "step": 1800
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.00019972778519732436,
-      "loss": 0.7528,
-      "step": 1900
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 0.0001996982322195061,
-      "loss": 0.725,
-      "step": 2000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.7452704310417175,
-      "eval_runtime": 1787.7554,
-      "eval_samples_per_second": 0.303,
-      "eval_steps_per_second": 0.303,
-      "step": 2000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00019966715903665795,
-      "loss": 0.7234,
-      "step": 2100
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 0.00019963456612258576,
-      "loss": 0.754,
-      "step": 2200
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00019960045397426841,
-      "loss": 0.7856,
-      "step": 2300
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00019956482311185006,
-      "loss": 0.7387,
-      "step": 2400
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 0.00019952767407863245,
-      "loss": 0.7309,
-      "step": 2500
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00019948900744106633,
-      "loss": 0.7232,
-      "step": 2600
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.00019944882378874316,
-      "loss": 0.7406,
-      "step": 2700
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 0.0001994071237343858,
-      "loss": 0.7166,
-      "step": 2800
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00019936390791383936,
-      "loss": 0.7308,
-      "step": 2900
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 0.00019931917698606143,
-      "loss": 0.7288,
-      "step": 3000
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 0.7343490123748779,
-      "eval_runtime": 1770.9966,
-      "eval_samples_per_second": 0.306,
-      "eval_steps_per_second": 0.306,
-      "step": 3000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00019927293163311206,
-      "loss": 0.7236,
-      "step": 3100
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00019922517256014337,
-      "loss": 0.716,
-      "step": 3200
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 0.00019917590049538874,
-      "loss": 0.7564,
-      "step": 3300
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00019912511619015177,
-      "loss": 0.7082,
-      "step": 3400
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00019907282041879484,
-      "loss": 0.7103,
-      "step": 3500
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 0.00019901901397872715,
-      "loss": 0.7457,
-      "step": 3600
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.0001989636976903928,
-      "loss": 0.7076,
-      "step": 3700
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 0.0001989068723972581,
-      "loss": 0.7217,
-      "step": 3800
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00019884853896579873,
-      "loss": 0.7175,
-      "step": 3900
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.0001987886982854866,
-      "loss": 0.7083,
-      "step": 4000
-    },
-    {
-      "epoch": 0.15,
-      "eval_loss": 0.726176917552948,
-      "eval_runtime": 1765.3933,
-      "eval_samples_per_second": 0.307,
-      "eval_steps_per_second": 0.307,
-      "step": 4000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 0.00019872735126877622,
-      "loss": 0.7228,
-      "step": 4100
-    },
-    {
-      "epoch": 0.16,
-      "learning_rate": 0.0001986644988510909,
-      "loss": 0.7133,
-      "step": 4200
-    },
-    {
-      "epoch": 0.16,
-      "learning_rate": 0.00019860014199080822,
-      "loss": 0.7243,
-      "step": 4300
-    },
-    {
-      "epoch": 0.16,
-      "learning_rate": 0.00019853428166924576,
-      "loss": 0.6929,
-      "step": 4400
-    },
-    {
-      "epoch": 0.17,
-      "learning_rate": 0.00019846691889064593,
-      "loss": 0.7392,
-      "step": 4500
-    },
-    {
-      "epoch": 0.17,
-      "learning_rate": 0.0001983980546821607,
-      "loss": 0.7247,
-      "step": 4600
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 0.0001983276900938359,
-      "loss": 0.7258,
-      "step": 4700
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 0.00019825582619859532,
-      "loss": 0.7197,
-      "step": 4800
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 0.0001981824640922242,
-      "loss": 0.6906,
-      "step": 4900
-    },
-    {
-      "epoch": 0.19,
-      "learning_rate": 0.00019810760489335266,
-      "loss": 0.7274,
-      "step": 5000
-    },
-    {
-      "epoch": 0.19,
-      "eval_loss": 0.7171670794487,
-      "eval_runtime": 1812.7597,
-      "eval_samples_per_second": 0.299,
-      "eval_steps_per_second": 0.299,
-      "step": 5000
-    },
-    {
-      "epoch": 0.19,
-      "learning_rate": 0.0001980312497434385,
-      "loss": 0.7105,
-      "step": 5100
-    },
-    {
-      "epoch": 0.19,
-      "learning_rate": 0.00019795339980675002,
-      "loss": 0.7091,
-      "step": 5200
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 0.00019787405627034804,
-      "loss": 0.7102,
-      "step": 5300
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 0.0001977932203440678,
-      "loss": 0.7314,
-      "step": 5400
-    },
-    {
-      "epoch": 0.21,
-      "learning_rate": 0.00019771089326050075,
-      "loss": 0.6945,
-      "step": 5500
-    },
-    {
-      "epoch": 0.21,
-      "learning_rate": 0.0001976270762749755,
-      "loss": 0.7048,
-      "step": 5600
-    },
-    {
-      "epoch": 0.21,
-      "learning_rate": 0.00019754177066553882,
-      "loss": 0.6963,
-      "step": 5700
-    },
-    {
-      "epoch": 0.22,
-      "learning_rate": 0.00019745497773293613,
-      "loss": 0.711,
-      "step": 5800
-    },
-    {
-      "epoch": 0.22,
-      "learning_rate": 0.0001973666988005916,
-      "loss": 0.7017,
-      "step": 5900
-    },
-    {
-      "epoch": 0.22,
-      "learning_rate": 0.00019727693521458806,
-      "loss": 0.7287,
-      "step": 6000
-    },
-    {
-      "epoch": 0.22,
-      "eval_loss": 0.710155725479126,
-      "eval_runtime": 1786.7467,
-      "eval_samples_per_second": 0.303,
-      "eval_steps_per_second": 0.303,
-      "step": 6000
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 0.00019718568834364638,
-      "loss": 0.6894,
-      "step": 6100
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 0.00019709295957910476,
-      "loss": 0.7061,
-      "step": 6200
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 0.00019699875033489728,
-      "loss": 0.7063,
-      "step": 6300
-    },
-    {
-      "epoch": 0.24,
-      "learning_rate": 0.00019690306204753254,
-      "loss": 0.6872,
-      "step": 6400
-    },
-    {
-      "epoch": 0.24,
-      "learning_rate": 0.0001968058961760717,
-      "loss": 0.7095,
-      "step": 6500
-    },
-    {
-      "epoch": 0.25,
-      "learning_rate": 0.00019670725420210618,
-      "loss": 0.695,
-      "step": 6600
-    },
-    {
-      "epoch": 0.25,
-      "learning_rate": 0.0001966071376297351,
-      "loss": 0.674,
-      "step": 6700
-    },
-    {
-      "epoch": 0.25,
-      "learning_rate": 0.00019650554798554236,
-      "loss": 0.7225,
-      "step": 6800
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 0.00019640248681857342,
-      "loss": 0.6845,
-      "step": 6900
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 0.00019629795570031149,
-      "loss": 0.6891,
-      "step": 7000
-    },
-    {
-      "epoch": 0.26,
-      "eval_loss": 0.703677773475647,
-      "eval_runtime": 1767.3593,
-      "eval_samples_per_second": 0.307,
-      "eval_steps_per_second": 0.307,
-      "step": 7000
-    }
-  ],
-  "max_steps": 80463,
-  "num_train_epochs": 3,
-  "total_flos": 1.962228951137157e+18,
-  "trial_name": null,
-  "trial_params": null
-}

checkpoint-7000/training_args.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9a281372a257be5e7ebbea8ac16ec51e707b224a5ed57c48e2f69549c5a031d8
-size 4027