diff --git a/codet5+/q90cce/checkpoint-16161/model.safetensors b/codet5+/q90cce/checkpoint-16161/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e10d4ae0c724236f3bf82d577a497d992585d88b --- /dev/null +++ b/codet5+/q90cce/checkpoint-16161/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bac365536c3abbcd19ea10781413295f880335e19e5faacc2a80cf87f759c4f3 +size 891558696 diff --git a/codet5+/q90cce/checkpoint-16161/optimizer.pt b/codet5+/q90cce/checkpoint-16161/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..a2752914b139998faf6f7f42fda152419ce655e3 --- /dev/null +++ b/codet5+/q90cce/checkpoint-16161/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8e947fc1a62608cdbd2e727ebf1b77ab2441741b82c210915feaa31490f802d2 +size 1783272762 diff --git a/codet5+/q90cce/checkpoint-16161/rng_state.pth b/codet5+/q90cce/checkpoint-16161/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..83982864b00f43f2679f8b268f7c4ec7bc586730 --- /dev/null +++ b/codet5+/q90cce/checkpoint-16161/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da7ab3d70319eedee0e20df99e3337c596300cf85c06a0ff43a497dc735bacc1 +size 14244 diff --git a/codet5+/q90cce/checkpoint-16161/scheduler.pt b/codet5+/q90cce/checkpoint-16161/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2ed7d1fe61cdd8ed4ae42c282293924f76b52287 --- /dev/null +++ b/codet5+/q90cce/checkpoint-16161/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cc6074681aa311442c0e33cfb9d6fdc6bf20b16d03df5498cbd742101c5b5b60 +size 1064 diff --git a/codet5+/q90cce/checkpoint-16161/training_args.bin b/codet5+/q90cce/checkpoint-16161/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..409dca7c0cf853a8cfd26d9c37b8482ada87cff7 --- /dev/null +++ b/codet5+/q90cce/checkpoint-16161/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bba5554b9a249566c939cc47d94f0779adfb4a53f5215181d0d81754796270a0 +size 4856 diff --git a/codet5+/q90cce/final_checkpoint/model.safetensors b/codet5+/q90cce/final_checkpoint/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e10d4ae0c724236f3bf82d577a497d992585d88b --- /dev/null +++ b/codet5+/q90cce/final_checkpoint/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bac365536c3abbcd19ea10781413295f880335e19e5faacc2a80cf87f759c4f3 +size 891558696 diff --git a/codet5+/q90useseq/checkpoint-16161/model.safetensors b/codet5+/q90useseq/checkpoint-16161/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..823164c40473f052f260216e8b26e478f835fe8c --- /dev/null +++ b/codet5+/q90useseq/checkpoint-16161/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9ef3dc2c3b199d6303258dcf37018964d5bdbc6ee10174a0a2c22597cee9f407 +size 891558696 diff --git a/codet5+/q90useseq/checkpoint-16161/optimizer.pt b/codet5+/q90useseq/checkpoint-16161/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..63972067f06eadf28617f1d0b8804b742445c9a4 --- /dev/null +++ b/codet5+/q90useseq/checkpoint-16161/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ac9866e8e98bd91dbfa964a33d1bb24d4f288b67351da114cf008c1f2a0a2a0f +size 1783272762 diff --git a/codet5+/q90useseq/checkpoint-16161/rng_state.pth b/codet5+/q90useseq/checkpoint-16161/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..2c9bda18f304573f76bc1de807630954fb471cd8 --- /dev/null +++ b/codet5+/q90useseq/checkpoint-16161/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bb2897a897fea1e576d8200facf0fd7c6684a305f296fe30fe6c438815665c88 +size 14244 diff --git a/codet5+/q90useseq/checkpoint-16161/scheduler.pt b/codet5+/q90useseq/checkpoint-16161/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..2ed7d1fe61cdd8ed4ae42c282293924f76b52287 --- /dev/null +++ b/codet5+/q90useseq/checkpoint-16161/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cc6074681aa311442c0e33cfb9d6fdc6bf20b16d03df5498cbd742101c5b5b60 +size 1064 diff --git a/codet5+/q90useseq/checkpoint-16161/training_args.bin b/codet5+/q90useseq/checkpoint-16161/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..5f4e37b56e9b288ccf639536388f3ceda4c21a80 --- /dev/null +++ b/codet5+/q90useseq/checkpoint-16161/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b9310a6cee128577aa0bc565fc71ba9d56a1ecc0f3386594477591138294a4c +size 4856 diff --git a/codet5+/q90useseq/events.out.tfevents.1712092321.nublar.444200.0 b/codet5+/q90useseq/events.out.tfevents.1712092321.nublar.444200.0 new file mode 100644 index 0000000000000000000000000000000000000000..3dfac18170476adf626d1dc690f301137e6f15ac --- /dev/null +++ b/codet5+/q90useseq/events.out.tfevents.1712092321.nublar.444200.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:64f61fc9afd31fc598f005fd9849bed2e87e11e4bc8944bf58e45f462a482386 +size 346166 diff --git a/codet5+/q90useseq/final_checkpoint/model.safetensors b/codet5+/q90useseq/final_checkpoint/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..823164c40473f052f260216e8b26e478f835fe8c --- /dev/null +++ b/codet5+/q90useseq/final_checkpoint/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9ef3dc2c3b199d6303258dcf37018964d5bdbc6ee10174a0a2c22597cee9f407 +size 891558696 diff --git a/gpt2/cgpt/out-cgpt-e18-cce/ckpt.pt b/gpt2/cgpt/out-cgpt-e18-cce/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..6a16d7cde57b3036f78ebdf0642b73633a7aa0b1 --- /dev/null +++ b/gpt2/cgpt/out-cgpt-e18-cce/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:84f04aedf6ad2067a3e0f3c07e6baeb77fd44f054169d95b8f137bce69c24ad7 +size 1493434662 diff --git a/gpt2/cgpt/out-cgpt-e18-use-seq-0.6/ckpt.pt b/gpt2/cgpt/out-cgpt-e18-use-seq-0.6/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..3f3691c8102ef6d61fc83ca291297b8166dffa0c --- /dev/null +++ b/gpt2/cgpt/out-cgpt-e18-use-seq-0.6/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1f1d6639eacd08e3920acd816dbf496bbb2c9c6db0b7fb296a014302dbb5daca +size 1493434662 diff --git a/gpt2/cgpt/out-cgpt-e18-use-seq-0.8/ckpt.pt b/gpt2/cgpt/out-cgpt-e18-use-seq-0.8/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..ead15abb6628f81ea4da879f6c3873c6c52b5928 --- /dev/null +++ b/gpt2/cgpt/out-cgpt-e18-use-seq-0.8/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eb6ae0ea207a5c0e6d630c087f01005a79aefb174ca9d212dbd5378d46bd0c8a +size 1493434662 diff --git a/gpt2/cgpt/out-cgpt-e18-use-seq-1.0/ckpt.pt b/gpt2/cgpt/out-cgpt-e18-use-seq-1.0/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..d117ec2c60de4c34a2580a8d505809cbe9f565a9 --- /dev/null +++ b/gpt2/cgpt/out-cgpt-e18-use-seq-1.0/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7bb276b90c587531cdb723c99cab892a84e8246d8e2969fe78a6c42ca1fd9b41 +size 1493434662 diff --git a/gpt2/cgpt/out-cgpt-e18-use-seq-1.2/ckpt.pt b/gpt2/cgpt/out-cgpt-e18-use-seq-1.2/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..8da90498ee5cd14c855792c8fe508ed6d959eb58 --- /dev/null +++ b/gpt2/cgpt/out-cgpt-e18-use-seq-1.2/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c7ebc9fb2d7be656a0be6c0a9fa495df645a474e87cd93320a15feaee58e94d6 +size 1493434662 diff --git a/gpt2/cgpt/out-cgpt-e18-use-seq-ablate-1.0/ckpt.pt b/gpt2/cgpt/out-cgpt-e18-use-seq-ablate-1.0/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..d884023f567ceb823cfd2281aefc2fa6aa4e6426 --- /dev/null +++ b/gpt2/cgpt/out-cgpt-e18-use-seq-ablate-1.0/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:03044a2681785ab3eef0d4ec81fc7a0ef0776ef063a4f0ace861f1125e0facc4 +size 1493434662 diff --git a/gpt2/q90/out-funcom-gpt2-e18-cce/ckpt.pt b/gpt2/q90/out-funcom-gpt2-e18-cce/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..a5cbc094cf9a01ceacb03094d184e129ecbe47b9 --- /dev/null +++ b/gpt2/q90/out-funcom-gpt2-e18-cce/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d34914d55af32caaa2156640abe6500732a9311532335588ebae1a81144baddb +size 1493434662 diff --git a/gpt2/q90/out-funcom-gpt2-e18-use-seq-0.6/ckpt.pt b/gpt2/q90/out-funcom-gpt2-e18-use-seq-0.6/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..163f719100a91b923785a4c224a9cb0eef9de361 --- /dev/null +++ b/gpt2/q90/out-funcom-gpt2-e18-use-seq-0.6/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ae52b73e6cc1472bdbe76caa18f81cf5ce5338c3c8c46a1666f67813265aa206 +size 1493434662 diff --git a/gpt2/q90/out-funcom-gpt2-e18-use-seq-0.8/ckpt.pt b/gpt2/q90/out-funcom-gpt2-e18-use-seq-0.8/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..8f699e765da222aa64b41dd1b435d5c6bd172351 --- /dev/null +++ b/gpt2/q90/out-funcom-gpt2-e18-use-seq-0.8/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e3bcd657d03fe534fd37a5ab0b82eae3a513c01b1ac1988209b7c0718b3bb2e0 +size 1493434662 diff --git a/gpt2/q90/out-funcom-gpt2-e18-use-seq-1.0/ckpt.pt b/gpt2/q90/out-funcom-gpt2-e18-use-seq-1.0/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..60cd9215cc49fba67d0a7bd2e63d180623daae77 --- /dev/null +++ b/gpt2/q90/out-funcom-gpt2-e18-use-seq-1.0/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fd5b45fd34b07dd6666f92d28f7b029d9e49e99ee2c825a4d0e5464e95575b33 +size 1493434662 diff --git a/gpt2/q90/out-funcom-gpt2-e18-use-seq-1.2/ckpt.pt b/gpt2/q90/out-funcom-gpt2-e18-use-seq-1.2/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..fe4352bb9aa9cb6eeb7d09a922f24203d52e3e20 --- /dev/null +++ b/gpt2/q90/out-funcom-gpt2-e18-use-seq-1.2/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bb141dd6e585b610594151369d078577c29a0d3d41957a359d3b37f4b7325e91 +size 1493434662 diff --git a/gpt2/q90/out-funcom-gpt2-e18-use-seq-ablate-1.0/ckpt.pt b/gpt2/q90/out-funcom-gpt2-e18-use-seq-ablate-1.0/ckpt.pt new file mode 100644 index 0000000000000000000000000000000000000000..181e9a0c17e98688bd2813d6c6e05d2d0fbff60b --- /dev/null +++ b/gpt2/q90/out-funcom-gpt2-e18-use-seq-ablate-1.0/ckpt.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cadab6f190a1293e6788681102e59de873048abf202fa6739448e96929ce432e +size 1493434726 diff --git a/llama/lora-alpaca-q90-cce/adapter_config.json b/llama/lora-alpaca-q90-cce/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..aeb529678e2f5a4463eb51f5ae04ecbbb9ede5bb --- /dev/null +++ b/llama/lora-alpaca-q90-cce/adapter_config.json @@ -0,0 +1,18 @@ +{ + "base_model_name_or_path": "decapoda-research/llama-7b-hf", + "bias": "none", + "enable_lora": null, + "fan_in_fan_out": false, + "inference_mode": true, + "lora_alpha": 16, + "lora_dropout": 0.05, + "merge_weights": false, + "modules_to_save": null, + "peft_type": "LORA", + "r": 8, + "target_modules": [ + "q_proj", + "v_proj" + ], + "task_type": "CAUSAL_LM" +} \ No newline at end of file diff --git a/llama/lora-alpaca-q90-cce/adapter_model.bin b/llama/lora-alpaca-q90-cce/adapter_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..aa10bc7a779ff74085263a87dd08914c83e8c2bf --- /dev/null +++ b/llama/lora-alpaca-q90-cce/adapter_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:85db8e950ef5d0f478d9511111d467d336e9f3b1a5951e1504bc0b097f42afa9 +size 16822989 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1000/optimizer.pt b/llama/lora-alpaca-q90-cce/checkpoint-1000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..fbaf1a3382db478a24d80b05680d9138709e0e7b --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:08448be733fb5513f2235cc506bbf6d8a3eed82d58e562bcd403a3187fb7d9a7 +size 33629893 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1000/pytorch_model.bin b/llama/lora-alpaca-q90-cce/checkpoint-1000/pytorch_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..1ec879eabcbc53156e6b365e299fd3dd43ab9dd5 --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1000/pytorch_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e5b1f8b85708871cade3d0285562fbb25d1ebe8040017504c66c30426c67e847 +size 16822989 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1000/rng_state.pth b/llama/lora-alpaca-q90-cce/checkpoint-1000/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..0adbc327ce8c428d417e7491bcf1ea5372748f3d --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1000/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:716b9e826437be7256498ff03ae610de855359aa3dcbc686fe327448af3b47ce +size 14575 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1000/scaler.pt b/llama/lora-alpaca-q90-cce/checkpoint-1000/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..efdbd3c795f6b0d4144e68355e99c220ccdedd09 --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1000/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:68cff80b680ddf6e7abbef98b5f336b97f9b5963e2209307f639383870e8cc71 +size 557 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1000/scheduler.pt b/llama/lora-alpaca-q90-cce/checkpoint-1000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..278bbb80daa6070e3daf26a13ea02afda68b9fbf --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a69c02d22e7ba13517e1bdfd0b1c3ab7029bb5e809777e0aaafead5cf2bf0531 +size 627 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1000/trainer_state.json b/llama/lora-alpaca-q90-cce/checkpoint-1000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..a7aaa5ed42adf16c36410553767b87896f8e461f --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1000/trainer_state.json @@ -0,0 +1,356 @@ +{ + "best_metric": 0.6784626245498657, + "best_model_checkpoint": "lora-alpaca/checkpoint-1000", + "epoch": 0.7511913425197775, + "global_step": 1000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.02, + "learning_rate": 5.9999999999999995e-05, + "loss": 2.2234, + "step": 20 + }, + { + "epoch": 0.03, + "learning_rate": 0.00011999999999999999, + "loss": 1.6986, + "step": 40 + }, + { + "epoch": 0.05, + "learning_rate": 0.00017999999999999998, + "loss": 1.0108, + "step": 60 + }, + { + "epoch": 0.06, + "learning_rate": 0.00023999999999999998, + "loss": 0.8498, + "step": 80 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 0.7975, + "step": 100 + }, + { + "epoch": 0.09, + "learning_rate": 0.00029512591389114535, + "loss": 0.7621, + "step": 120 + }, + { + "epoch": 0.11, + "learning_rate": 0.0002902518277822908, + "loss": 0.7465, + "step": 140 + }, + { + "epoch": 0.12, + "learning_rate": 0.0002853777416734362, + "loss": 0.7367, + "step": 160 + }, + { + "epoch": 0.14, + "learning_rate": 0.0002805036555645816, + "loss": 0.732, + "step": 180 + }, + { + "epoch": 0.15, + "learning_rate": 0.00027562956945572704, + "loss": 0.7253, + "step": 200 + }, + { + "epoch": 0.15, + "eval_loss": 0.7244793176651001, + "eval_runtime": 129.0223, + "eval_samples_per_second": 15.501, + "eval_steps_per_second": 1.938, + "step": 200 + }, + { + "epoch": 0.17, + "learning_rate": 0.0002707554833468724, + "loss": 0.7226, + "step": 220 + }, + { + "epoch": 0.18, + "learning_rate": 0.00026588139723801785, + "loss": 0.716, + "step": 240 + }, + { + "epoch": 0.2, + "learning_rate": 0.00026100731112916323, + "loss": 0.7182, + "step": 260 + }, + { + "epoch": 0.21, + "learning_rate": 0.00025613322502030867, + "loss": 0.7026, + "step": 280 + }, + { + "epoch": 0.23, + "learning_rate": 0.0002512591389114541, + "loss": 0.7139, + "step": 300 + }, + { + "epoch": 0.24, + "learning_rate": 0.0002463850528025995, + "loss": 0.7156, + "step": 320 + }, + { + "epoch": 0.26, + "learning_rate": 0.00024151096669374492, + "loss": 0.7069, + "step": 340 + }, + { + "epoch": 0.27, + "learning_rate": 0.0002366368805848903, + "loss": 0.6998, + "step": 360 + }, + { + "epoch": 0.29, + "learning_rate": 0.0002317627944760357, + "loss": 0.7058, + "step": 380 + }, + { + "epoch": 0.3, + "learning_rate": 0.00022688870836718114, + "loss": 0.7004, + "step": 400 + }, + { + "epoch": 0.3, + "eval_loss": 0.6993061900138855, + "eval_runtime": 129.0376, + "eval_samples_per_second": 15.499, + "eval_steps_per_second": 1.937, + "step": 400 + }, + { + "epoch": 0.32, + "learning_rate": 0.00022201462225832652, + "loss": 0.6933, + "step": 420 + }, + { + "epoch": 0.33, + "learning_rate": 0.00021714053614947196, + "loss": 0.6868, + "step": 440 + }, + { + "epoch": 0.35, + "learning_rate": 0.00021226645004061737, + "loss": 0.6955, + "step": 460 + }, + { + "epoch": 0.36, + "learning_rate": 0.00020739236393176277, + "loss": 0.6878, + "step": 480 + }, + { + "epoch": 0.38, + "learning_rate": 0.00020251827782290818, + "loss": 0.6891, + "step": 500 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001976441917140536, + "loss": 0.686, + "step": 520 + }, + { + "epoch": 0.41, + "learning_rate": 0.00019277010560519902, + "loss": 0.6977, + "step": 540 + }, + { + "epoch": 0.42, + "learning_rate": 0.00018789601949634443, + "loss": 0.688, + "step": 560 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001830219333874898, + "loss": 0.6838, + "step": 580 + }, + { + "epoch": 0.45, + "learning_rate": 0.00017814784727863525, + "loss": 0.682, + "step": 600 + }, + { + "epoch": 0.45, + "eval_loss": 0.6886340379714966, + "eval_runtime": 128.9722, + "eval_samples_per_second": 15.507, + "eval_steps_per_second": 1.938, + "step": 600 + }, + { + "epoch": 0.47, + "learning_rate": 0.00017327376116978065, + "loss": 0.6879, + "step": 620 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001683996750609261, + "loss": 0.6871, + "step": 640 + }, + { + "epoch": 0.5, + "learning_rate": 0.00016352558895207147, + "loss": 0.6823, + "step": 660 + }, + { + "epoch": 0.51, + "learning_rate": 0.00015865150284321688, + "loss": 0.6732, + "step": 680 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001537774167343623, + "loss": 0.6832, + "step": 700 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001489033306255077, + "loss": 0.6795, + "step": 720 + }, + { + "epoch": 0.56, + "learning_rate": 0.00014402924451665313, + "loss": 0.686, + "step": 740 + }, + { + "epoch": 0.57, + "learning_rate": 0.00013915515840779853, + "loss": 0.6799, + "step": 760 + }, + { + "epoch": 0.59, + "learning_rate": 0.00013428107229894394, + "loss": 0.6738, + "step": 780 + }, + { + "epoch": 0.6, + "learning_rate": 0.00012940698619008935, + "loss": 0.6801, + "step": 800 + }, + { + "epoch": 0.6, + "eval_loss": 0.6822482943534851, + "eval_runtime": 128.9861, + "eval_samples_per_second": 15.506, + "eval_steps_per_second": 1.938, + "step": 800 + }, + { + "epoch": 0.62, + "learning_rate": 0.00012453290008123476, + "loss": 0.6803, + "step": 820 + }, + { + "epoch": 0.63, + "learning_rate": 0.00011965881397238017, + "loss": 0.6647, + "step": 840 + }, + { + "epoch": 0.65, + "learning_rate": 0.00011478472786352559, + "loss": 0.6838, + "step": 860 + }, + { + "epoch": 0.66, + "learning_rate": 0.000109910641754671, + "loss": 0.6793, + "step": 880 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001050365556458164, + "loss": 0.6882, + "step": 900 + }, + { + "epoch": 0.69, + "learning_rate": 0.00010016246953696181, + "loss": 0.6733, + "step": 920 + }, + { + "epoch": 0.71, + "learning_rate": 9.528838342810722e-05, + "loss": 0.6786, + "step": 940 + }, + { + "epoch": 0.72, + "learning_rate": 9.041429731925264e-05, + "loss": 0.682, + "step": 960 + }, + { + "epoch": 0.74, + "learning_rate": 8.554021121039803e-05, + "loss": 0.6737, + "step": 980 + }, + { + "epoch": 0.75, + "learning_rate": 8.066612510154345e-05, + "loss": 0.6681, + "step": 1000 + }, + { + "epoch": 0.75, + "eval_loss": 0.6784626245498657, + "eval_runtime": 129.063, + "eval_samples_per_second": 15.496, + "eval_steps_per_second": 1.937, + "step": 1000 + } + ], + "max_steps": 1331, + "num_train_epochs": 1, + "total_flos": 1.29988124737536e+18, + "trial_name": null, + "trial_params": null +} diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1000/training_args.bin b/llama/lora-alpaca-q90-cce/checkpoint-1000/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..a339d1798fdc6ddbd10b0635a433166e8f42c435 --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1000/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fac929355b6e3b85d753318e1aa77cfc30f00e526d22857f9fad037d422dcc46 +size 3515 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1200/optimizer.pt b/llama/lora-alpaca-q90-cce/checkpoint-1200/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..446e67b9b4796d97e7bbb2c37a2a446090078d4d --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1200/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:33c49815930ea90b0dac6ef1171f469dc470d3cdcb9c57f454da076b9a4f9ab8 +size 33629893 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1200/pytorch_model.bin b/llama/lora-alpaca-q90-cce/checkpoint-1200/pytorch_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..419f11682f4b16945aa3bde68c36ed1a1434728f --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1200/pytorch_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ba9f60fb87ca30385c9a4917ba3ebeedc71c97e0a283b0324e79a1b8ce328ba7 +size 16822989 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1200/rng_state.pth b/llama/lora-alpaca-q90-cce/checkpoint-1200/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..77a27d6412fdbe790e86d3f757db3b95b6efe363 --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1200/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f4d98aaa8ee8a726667d1326e2de60a7ee06aaece6c618bc7ce732b91f5e2f50 +size 14575 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1200/scaler.pt b/llama/lora-alpaca-q90-cce/checkpoint-1200/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..c73b6e7148d8ae7026711173634e0a11b1b94e2d --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1200/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:741cefeca9ef427f92406d2d10b81996655e2a9d50eb7aaa9614e6fdd1c9f529 +size 557 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1200/scheduler.pt b/llama/lora-alpaca-q90-cce/checkpoint-1200/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..8b94a6c830e58ac4dd247e6874c21c29273368e2 --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1200/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4d90a20c54899e27f19a89e3464bfe4b304595c9ee6541365c21acdedbea9c09 +size 627 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1200/trainer_state.json b/llama/lora-alpaca-q90-cce/checkpoint-1200/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..f2bffc01b478a9f4b2e5c2d1c3bccc42dd2db70f --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1200/trainer_state.json @@ -0,0 +1,424 @@ +{ + "best_metric": 0.6754332184791565, + "best_model_checkpoint": "lora-alpaca/checkpoint-1200", + "epoch": 0.901429611023733, + "global_step": 1200, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.02, + "learning_rate": 5.9999999999999995e-05, + "loss": 2.2234, + "step": 20 + }, + { + "epoch": 0.03, + "learning_rate": 0.00011999999999999999, + "loss": 1.6986, + "step": 40 + }, + { + "epoch": 0.05, + "learning_rate": 0.00017999999999999998, + "loss": 1.0108, + "step": 60 + }, + { + "epoch": 0.06, + "learning_rate": 0.00023999999999999998, + "loss": 0.8498, + "step": 80 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 0.7975, + "step": 100 + }, + { + "epoch": 0.09, + "learning_rate": 0.00029512591389114535, + "loss": 0.7621, + "step": 120 + }, + { + "epoch": 0.11, + "learning_rate": 0.0002902518277822908, + "loss": 0.7465, + "step": 140 + }, + { + "epoch": 0.12, + "learning_rate": 0.0002853777416734362, + "loss": 0.7367, + "step": 160 + }, + { + "epoch": 0.14, + "learning_rate": 0.0002805036555645816, + "loss": 0.732, + "step": 180 + }, + { + "epoch": 0.15, + "learning_rate": 0.00027562956945572704, + "loss": 0.7253, + "step": 200 + }, + { + "epoch": 0.15, + "eval_loss": 0.7244793176651001, + "eval_runtime": 129.0223, + "eval_samples_per_second": 15.501, + "eval_steps_per_second": 1.938, + "step": 200 + }, + { + "epoch": 0.17, + "learning_rate": 0.0002707554833468724, + "loss": 0.7226, + "step": 220 + }, + { + "epoch": 0.18, + "learning_rate": 0.00026588139723801785, + "loss": 0.716, + "step": 240 + }, + { + "epoch": 0.2, + "learning_rate": 0.00026100731112916323, + "loss": 0.7182, + "step": 260 + }, + { + "epoch": 0.21, + "learning_rate": 0.00025613322502030867, + "loss": 0.7026, + "step": 280 + }, + { + "epoch": 0.23, + "learning_rate": 0.0002512591389114541, + "loss": 0.7139, + "step": 300 + }, + { + "epoch": 0.24, + "learning_rate": 0.0002463850528025995, + "loss": 0.7156, + "step": 320 + }, + { + "epoch": 0.26, + "learning_rate": 0.00024151096669374492, + "loss": 0.7069, + "step": 340 + }, + { + "epoch": 0.27, + "learning_rate": 0.0002366368805848903, + "loss": 0.6998, + "step": 360 + }, + { + "epoch": 0.29, + "learning_rate": 0.0002317627944760357, + "loss": 0.7058, + "step": 380 + }, + { + "epoch": 0.3, + "learning_rate": 0.00022688870836718114, + "loss": 0.7004, + "step": 400 + }, + { + "epoch": 0.3, + "eval_loss": 0.6993061900138855, + "eval_runtime": 129.0376, + "eval_samples_per_second": 15.499, + "eval_steps_per_second": 1.937, + "step": 400 + }, + { + "epoch": 0.32, + "learning_rate": 0.00022201462225832652, + "loss": 0.6933, + "step": 420 + }, + { + "epoch": 0.33, + "learning_rate": 0.00021714053614947196, + "loss": 0.6868, + "step": 440 + }, + { + "epoch": 0.35, + "learning_rate": 0.00021226645004061737, + "loss": 0.6955, + "step": 460 + }, + { + "epoch": 0.36, + "learning_rate": 0.00020739236393176277, + "loss": 0.6878, + "step": 480 + }, + { + "epoch": 0.38, + "learning_rate": 0.00020251827782290818, + "loss": 0.6891, + "step": 500 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001976441917140536, + "loss": 0.686, + "step": 520 + }, + { + "epoch": 0.41, + "learning_rate": 0.00019277010560519902, + "loss": 0.6977, + "step": 540 + }, + { + "epoch": 0.42, + "learning_rate": 0.00018789601949634443, + "loss": 0.688, + "step": 560 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001830219333874898, + "loss": 0.6838, + "step": 580 + }, + { + "epoch": 0.45, + "learning_rate": 0.00017814784727863525, + "loss": 0.682, + "step": 600 + }, + { + "epoch": 0.45, + "eval_loss": 0.6886340379714966, + "eval_runtime": 128.9722, + "eval_samples_per_second": 15.507, + "eval_steps_per_second": 1.938, + "step": 600 + }, + { + "epoch": 0.47, + "learning_rate": 0.00017327376116978065, + "loss": 0.6879, + "step": 620 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001683996750609261, + "loss": 0.6871, + "step": 640 + }, + { + "epoch": 0.5, + "learning_rate": 0.00016352558895207147, + "loss": 0.6823, + "step": 660 + }, + { + "epoch": 0.51, + "learning_rate": 0.00015865150284321688, + "loss": 0.6732, + "step": 680 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001537774167343623, + "loss": 0.6832, + "step": 700 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001489033306255077, + "loss": 0.6795, + "step": 720 + }, + { + "epoch": 0.56, + "learning_rate": 0.00014402924451665313, + "loss": 0.686, + "step": 740 + }, + { + "epoch": 0.57, + "learning_rate": 0.00013915515840779853, + "loss": 0.6799, + "step": 760 + }, + { + "epoch": 0.59, + "learning_rate": 0.00013428107229894394, + "loss": 0.6738, + "step": 780 + }, + { + "epoch": 0.6, + "learning_rate": 0.00012940698619008935, + "loss": 0.6801, + "step": 800 + }, + { + "epoch": 0.6, + "eval_loss": 0.6822482943534851, + "eval_runtime": 128.9861, + "eval_samples_per_second": 15.506, + "eval_steps_per_second": 1.938, + "step": 800 + }, + { + "epoch": 0.62, + "learning_rate": 0.00012453290008123476, + "loss": 0.6803, + "step": 820 + }, + { + "epoch": 0.63, + "learning_rate": 0.00011965881397238017, + "loss": 0.6647, + "step": 840 + }, + { + "epoch": 0.65, + "learning_rate": 0.00011478472786352559, + "loss": 0.6838, + "step": 860 + }, + { + "epoch": 0.66, + "learning_rate": 0.000109910641754671, + "loss": 0.6793, + "step": 880 + }, + { + "epoch": 0.68, + "learning_rate": 0.0001050365556458164, + "loss": 0.6882, + "step": 900 + }, + { + "epoch": 0.69, + "learning_rate": 0.00010016246953696181, + "loss": 0.6733, + "step": 920 + }, + { + "epoch": 0.71, + "learning_rate": 9.528838342810722e-05, + "loss": 0.6786, + "step": 940 + }, + { + "epoch": 0.72, + "learning_rate": 9.041429731925264e-05, + "loss": 0.682, + "step": 960 + }, + { + "epoch": 0.74, + "learning_rate": 8.554021121039803e-05, + "loss": 0.6737, + "step": 980 + }, + { + "epoch": 0.75, + "learning_rate": 8.066612510154345e-05, + "loss": 0.6681, + "step": 1000 + }, + { + "epoch": 0.75, + "eval_loss": 0.6784626245498657, + "eval_runtime": 129.063, + "eval_samples_per_second": 15.496, + "eval_steps_per_second": 1.937, + "step": 1000 + }, + { + "epoch": 0.77, + "learning_rate": 7.579203899268886e-05, + "loss": 0.678, + "step": 1020 + }, + { + "epoch": 0.78, + "learning_rate": 7.091795288383428e-05, + "loss": 0.6692, + "step": 1040 + }, + { + "epoch": 0.8, + "learning_rate": 6.604386677497969e-05, + "loss": 0.6767, + "step": 1060 + }, + { + "epoch": 0.81, + "learning_rate": 6.11697806661251e-05, + "loss": 0.6699, + "step": 1080 + }, + { + "epoch": 0.83, + "learning_rate": 5.6295694557270505e-05, + "loss": 0.678, + "step": 1100 + }, + { + "epoch": 0.84, + "learning_rate": 5.142160844841592e-05, + "loss": 0.6802, + "step": 1120 + }, + { + "epoch": 0.86, + "learning_rate": 4.654752233956133e-05, + "loss": 0.6773, + "step": 1140 + }, + { + "epoch": 0.87, + "learning_rate": 4.167343623070674e-05, + "loss": 0.6719, + "step": 1160 + }, + { + "epoch": 0.89, + "learning_rate": 3.679935012185215e-05, + "loss": 0.6764, + "step": 1180 + }, + { + "epoch": 0.9, + "learning_rate": 3.192526401299756e-05, + "loss": 0.6745, + "step": 1200 + }, + { + "epoch": 0.9, + "eval_loss": 0.6754332184791565, + "eval_runtime": 128.9887, + "eval_samples_per_second": 15.505, + "eval_steps_per_second": 1.938, + "step": 1200 + } + ], + "max_steps": 1331, + "num_train_epochs": 1, + "total_flos": 1.559857496850432e+18, + "trial_name": null, + "trial_params": null +} diff --git a/llama/lora-alpaca-q90-cce/checkpoint-1200/training_args.bin b/llama/lora-alpaca-q90-cce/checkpoint-1200/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..a339d1798fdc6ddbd10b0635a433166e8f42c435 --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-1200/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fac929355b6e3b85d753318e1aa77cfc30f00e526d22857f9fad037d422dcc46 +size 3515 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-800/optimizer.pt b/llama/lora-alpaca-q90-cce/checkpoint-800/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..f8e3baa203292e3669821f557b3a122c535a49b3 --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-800/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4d4dae287198f854ec748229003488aa7600420a32ba91d46f4d84bd52120b57 +size 33629893 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-800/pytorch_model.bin b/llama/lora-alpaca-q90-cce/checkpoint-800/pytorch_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..17c81b20c5cb3983a898ae1403e68ccbcbc890a3 --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-800/pytorch_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:acd23c87bf974e5147bc9a152a9262c3efffaa61b736daf495c3cf4da9f237d6 +size 16822989 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-800/rng_state.pth b/llama/lora-alpaca-q90-cce/checkpoint-800/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..79837296b27095e3d8b9fa2cd4c7b636538ce7c3 --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-800/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f1a4cc8c5baf51f3cd71e0d56a3c0b65204e606c6f5eb0adc2d7ae2c7e900fbf +size 14575 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-800/scaler.pt b/llama/lora-alpaca-q90-cce/checkpoint-800/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..5e01dd7b5d3a8968bb4c73a805f08f0f65c9b57f --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-800/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:27ec07a12731ae6f9765d05fe7c8495505f1d0f90b4cc6255a0853fec3970808 +size 557 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-800/scheduler.pt b/llama/lora-alpaca-q90-cce/checkpoint-800/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..cc15c7ca0e6513cacd323637d92497583f0ca6cd --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-800/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:616a7b1df89e3ef38b6fa5a5f381236e37a81295f1e5fe4260a5014a64e9a421 +size 627 diff --git a/llama/lora-alpaca-q90-cce/checkpoint-800/trainer_state.json b/llama/lora-alpaca-q90-cce/checkpoint-800/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..15ecfd4e67880cf73c4942945ffa4110de7bdf22 --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-800/trainer_state.json @@ -0,0 +1,288 @@ +{ + "best_metric": 0.6822482943534851, + "best_model_checkpoint": "lora-alpaca/checkpoint-800", + "epoch": 0.600953074015822, + "global_step": 800, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.02, + "learning_rate": 5.9999999999999995e-05, + "loss": 2.2234, + "step": 20 + }, + { + "epoch": 0.03, + "learning_rate": 0.00011999999999999999, + "loss": 1.6986, + "step": 40 + }, + { + "epoch": 0.05, + "learning_rate": 0.00017999999999999998, + "loss": 1.0108, + "step": 60 + }, + { + "epoch": 0.06, + "learning_rate": 0.00023999999999999998, + "loss": 0.8498, + "step": 80 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 0.7975, + "step": 100 + }, + { + "epoch": 0.09, + "learning_rate": 0.00029512591389114535, + "loss": 0.7621, + "step": 120 + }, + { + "epoch": 0.11, + "learning_rate": 0.0002902518277822908, + "loss": 0.7465, + "step": 140 + }, + { + "epoch": 0.12, + "learning_rate": 0.0002853777416734362, + "loss": 0.7367, + "step": 160 + }, + { + "epoch": 0.14, + "learning_rate": 0.0002805036555645816, + "loss": 0.732, + "step": 180 + }, + { + "epoch": 0.15, + "learning_rate": 0.00027562956945572704, + "loss": 0.7253, + "step": 200 + }, + { + "epoch": 0.15, + "eval_loss": 0.7244793176651001, + "eval_runtime": 129.0223, + "eval_samples_per_second": 15.501, + "eval_steps_per_second": 1.938, + "step": 200 + }, + { + "epoch": 0.17, + "learning_rate": 0.0002707554833468724, + "loss": 0.7226, + "step": 220 + }, + { + "epoch": 0.18, + "learning_rate": 0.00026588139723801785, + "loss": 0.716, + "step": 240 + }, + { + "epoch": 0.2, + "learning_rate": 0.00026100731112916323, + "loss": 0.7182, + "step": 260 + }, + { + "epoch": 0.21, + "learning_rate": 0.00025613322502030867, + "loss": 0.7026, + "step": 280 + }, + { + "epoch": 0.23, + "learning_rate": 0.0002512591389114541, + "loss": 0.7139, + "step": 300 + }, + { + "epoch": 0.24, + "learning_rate": 0.0002463850528025995, + "loss": 0.7156, + "step": 320 + }, + { + "epoch": 0.26, + "learning_rate": 0.00024151096669374492, + "loss": 0.7069, + "step": 340 + }, + { + "epoch": 0.27, + "learning_rate": 0.0002366368805848903, + "loss": 0.6998, + "step": 360 + }, + { + "epoch": 0.29, + "learning_rate": 0.0002317627944760357, + "loss": 0.7058, + "step": 380 + }, + { + "epoch": 0.3, + "learning_rate": 0.00022688870836718114, + "loss": 0.7004, + "step": 400 + }, + { + "epoch": 0.3, + "eval_loss": 0.6993061900138855, + "eval_runtime": 129.0376, + "eval_samples_per_second": 15.499, + "eval_steps_per_second": 1.937, + "step": 400 + }, + { + "epoch": 0.32, + "learning_rate": 0.00022201462225832652, + "loss": 0.6933, + "step": 420 + }, + { + "epoch": 0.33, + "learning_rate": 0.00021714053614947196, + "loss": 0.6868, + "step": 440 + }, + { + "epoch": 0.35, + "learning_rate": 0.00021226645004061737, + "loss": 0.6955, + "step": 460 + }, + { + "epoch": 0.36, + "learning_rate": 0.00020739236393176277, + "loss": 0.6878, + "step": 480 + }, + { + "epoch": 0.38, + "learning_rate": 0.00020251827782290818, + "loss": 0.6891, + "step": 500 + }, + { + "epoch": 0.39, + "learning_rate": 0.0001976441917140536, + "loss": 0.686, + "step": 520 + }, + { + "epoch": 0.41, + "learning_rate": 0.00019277010560519902, + "loss": 0.6977, + "step": 540 + }, + { + "epoch": 0.42, + "learning_rate": 0.00018789601949634443, + "loss": 0.688, + "step": 560 + }, + { + "epoch": 0.44, + "learning_rate": 0.0001830219333874898, + "loss": 0.6838, + "step": 580 + }, + { + "epoch": 0.45, + "learning_rate": 0.00017814784727863525, + "loss": 0.682, + "step": 600 + }, + { + "epoch": 0.45, + "eval_loss": 0.6886340379714966, + "eval_runtime": 128.9722, + "eval_samples_per_second": 15.507, + "eval_steps_per_second": 1.938, + "step": 600 + }, + { + "epoch": 0.47, + "learning_rate": 0.00017327376116978065, + "loss": 0.6879, + "step": 620 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001683996750609261, + "loss": 0.6871, + "step": 640 + }, + { + "epoch": 0.5, + "learning_rate": 0.00016352558895207147, + "loss": 0.6823, + "step": 660 + }, + { + "epoch": 0.51, + "learning_rate": 0.00015865150284321688, + "loss": 0.6732, + "step": 680 + }, + { + "epoch": 0.53, + "learning_rate": 0.0001537774167343623, + "loss": 0.6832, + "step": 700 + }, + { + "epoch": 0.54, + "learning_rate": 0.0001489033306255077, + "loss": 0.6795, + "step": 720 + }, + { + "epoch": 0.56, + "learning_rate": 0.00014402924451665313, + "loss": 0.686, + "step": 740 + }, + { + "epoch": 0.57, + "learning_rate": 0.00013915515840779853, + "loss": 0.6799, + "step": 760 + }, + { + "epoch": 0.59, + "learning_rate": 0.00013428107229894394, + "loss": 0.6738, + "step": 780 + }, + { + "epoch": 0.6, + "learning_rate": 0.00012940698619008935, + "loss": 0.6801, + "step": 800 + }, + { + "epoch": 0.6, + "eval_loss": 0.6822482943534851, + "eval_runtime": 128.9861, + "eval_samples_per_second": 15.506, + "eval_steps_per_second": 1.938, + "step": 800 + } + ], + "max_steps": 1331, + "num_train_epochs": 1, + "total_flos": 1.039904997900288e+18, + "trial_name": null, + "trial_params": null +} diff --git a/llama/lora-alpaca-q90-cce/checkpoint-800/training_args.bin b/llama/lora-alpaca-q90-cce/checkpoint-800/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..a339d1798fdc6ddbd10b0635a433166e8f42c435 --- /dev/null +++ b/llama/lora-alpaca-q90-cce/checkpoint-800/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fac929355b6e3b85d753318e1aa77cfc30f00e526d22857f9fad037d422dcc46 +size 3515 diff --git a/llama/lora-alpaca-q90-use-seq/adapter_config.json b/llama/lora-alpaca-q90-use-seq/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..ab29193b9edba2f2fdf333d26a65a6f56838fe90 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/adapter_config.json @@ -0,0 +1,19 @@ +{ + "base_model_name_or_path": "decapoda-research/llama-7b-hf", + "bias": "none", + "enable_lora": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "lora_alpha": 16, + "lora_dropout": 0.05, + "merge_weights": false, + "modules_to_save": null, + "peft_type": "LORA", + "r": 8, + "target_modules": [ + "q_proj", + "v_proj" + ], + "task_type": "CAUSAL_LM" +} \ No newline at end of file diff --git a/llama/lora-alpaca-q90-use-seq/adapter_model.bin b/llama/lora-alpaca-q90-use-seq/adapter_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..a894b450a88c5fba149d110f1c2ba45d24e6820e --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/adapter_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4d4f45e8904cd0ecec21041e4968cee3f330f05b89c3aa425e3eb7d9a3251ec4 +size 16822989 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1070/optimizer.pt b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..38f71d7b754ed222fcbdb609538d5cc0d7855f38 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:457469443eca84dd2558b97703425a20dc9fd826a73cf259c6523b557887d08d +size 33629893 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1070/pytorch_model.bin b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/pytorch_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..8d8b514e6bd053604535d2c5f2c7c2d4ebb1952d --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/pytorch_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e6d0e3089f64bceb60773dd69b8e35d2042fb0389fca59c0e9f8bef6b63b4bc4 +size 16822989 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1070/rng_state.pth b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..228976bda55b2ee4492ebf9cdabb3a7bcd13817e --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fb90a72ee3c4c9903ac7f4d3896af0846d0138f4003aebaa211c12604a50afb0 +size 14575 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1070/scaler.pt b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..dfa5b2fc14890b67fa81ef8afddd02e4e2173e8b --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5260e5905b216ea38f63f19b666a9ebed81cc9a2f9b9fe9f54d232eee04f4692 +size 557 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1070/scheduler.pt b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..4bb157ec86c7cba094e55bed6f5c62da1a0ae50f --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c406c613803e729f1e0946f32317aad56134ee698208ff974b8835f01a71b20c +size 627 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1070/special_tokens_map.json b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..80e8b5de5e5d99c398943f1d5c9a51c3ade8c376 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + } +} diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1070/tokenizer.model b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..6c00c742ce03c627d6cd5b795984876fa49fa899 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347 +size 499723 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1070/tokenizer_config.json b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..381459f4531afd0797d7e2117b21541618a5ad58 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/tokenizer_config.json @@ -0,0 +1,34 @@ +{ + "add_bos_token": true, + "add_eos_token": true, + "bos_token": { + "__type": "AddedToken", + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "clean_up_tokenization_spaces": false, + "eos_token": { + "__type": "AddedToken", + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "model_max_length": 1000000000000000019884624838656, + "pad_token": null, + "sp_model_kwargs": {}, + "special_tokens_map_file": "/home/chiayi/.cache/huggingface/hub/models--decapoda-research--llama-7b-hf/snapshots/5f98eefcc80e437ef68d457ad7bf167c2c6a1348/special_tokens_map.json", + "tokenizer_class": "LlamaTokenizer", + "unk_token": { + "__type": "AddedToken", + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + } +} diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1070/trainer_state.json b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..a91ad5f88041fe918f0bb41e3217e03ade448f01 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/trainer_state.json @@ -0,0 +1,1190 @@ +{ + "best_metric": 0.7264513969421387, + "best_model_checkpoint": "lora-alpaca-use-seq/checkpoint-1070", + "epoch": 0.794449987238683, + "global_step": 1070, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.01, + "eval_loss": 2.4221816062927246, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.348, + "eval_steps_per_second": 6.348, + "step": 10 + }, + { + "epoch": 0.01, + "learning_rate": 5.9999999999999995e-05, + "loss": 2.4934, + "step": 20 + }, + { + "epoch": 0.01, + "eval_loss": 2.3228158950805664, + "eval_runtime": 0.1587, + "eval_samples_per_second": 6.302, + "eval_steps_per_second": 6.302, + "step": 20 + }, + { + "epoch": 0.02, + "eval_loss": 1.8552848100662231, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.386, + "eval_steps_per_second": 6.386, + "step": 30 + }, + { + "epoch": 0.03, + "learning_rate": 0.000117, + "loss": 1.9804, + "step": 40 + }, + { + "epoch": 0.03, + "eval_loss": 1.2810053825378418, + "eval_runtime": 0.1559, + "eval_samples_per_second": 6.416, + "eval_steps_per_second": 6.416, + "step": 40 + }, + { + "epoch": 0.04, + "eval_loss": 0.9789324998855591, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.382, + "eval_steps_per_second": 6.382, + "step": 50 + }, + { + "epoch": 0.04, + "learning_rate": 0.00017699999999999997, + "loss": 1.165, + "step": 60 + }, + { + "epoch": 0.04, + "eval_loss": 0.9216629266738892, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.343, + "eval_steps_per_second": 6.343, + "step": 60 + }, + { + "epoch": 0.05, + "eval_loss": 0.8872115612030029, + "eval_runtime": 0.1579, + "eval_samples_per_second": 6.333, + "eval_steps_per_second": 6.333, + "step": 70 + }, + { + "epoch": 0.06, + "learning_rate": 0.000237, + "loss": 0.9931, + "step": 80 + }, + { + "epoch": 0.06, + "eval_loss": 0.8584132194519043, + "eval_runtime": 0.1599, + "eval_samples_per_second": 6.254, + "eval_steps_per_second": 6.254, + "step": 80 + }, + { + "epoch": 0.07, + "eval_loss": 0.9114644527435303, + "eval_runtime": 0.1598, + "eval_samples_per_second": 6.259, + "eval_steps_per_second": 6.259, + "step": 90 + }, + { + "epoch": 0.07, + "learning_rate": 0.00029699999999999996, + "loss": 0.9545, + "step": 100 + }, + { + "epoch": 0.07, + "eval_loss": 0.8909071087837219, + "eval_runtime": 0.1603, + "eval_samples_per_second": 6.24, + "eval_steps_per_second": 6.24, + "step": 100 + }, + { + "epoch": 0.08, + "eval_loss": 0.9121037721633911, + "eval_runtime": 0.1598, + "eval_samples_per_second": 6.259, + "eval_steps_per_second": 6.259, + "step": 110 + }, + { + "epoch": 0.09, + "learning_rate": 0.0002954253611556982, + "loss": 0.8947, + "step": 120 + }, + { + "epoch": 0.09, + "eval_loss": 0.8340358734130859, + "eval_runtime": 0.1605, + "eval_samples_per_second": 6.229, + "eval_steps_per_second": 6.229, + "step": 120 + }, + { + "epoch": 0.1, + "eval_loss": 0.8849254846572876, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.33, + "eval_steps_per_second": 6.33, + "step": 130 + }, + { + "epoch": 0.1, + "learning_rate": 0.0002906099518459069, + "loss": 0.8711, + "step": 140 + }, + { + "epoch": 0.1, + "eval_loss": 0.8834758996963501, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.368, + "eval_steps_per_second": 6.368, + "step": 140 + }, + { + "epoch": 0.11, + "eval_loss": 0.8293902277946472, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.316, + "eval_steps_per_second": 6.316, + "step": 150 + }, + { + "epoch": 0.12, + "learning_rate": 0.00028579454253611556, + "loss": 0.8688, + "step": 160 + }, + { + "epoch": 0.12, + "eval_loss": 0.8478549718856812, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.324, + "eval_steps_per_second": 6.324, + "step": 160 + }, + { + "epoch": 0.13, + "eval_loss": 0.8488869667053223, + "eval_runtime": 0.159, + "eval_samples_per_second": 6.288, + "eval_steps_per_second": 6.288, + "step": 170 + }, + { + "epoch": 0.13, + "learning_rate": 0.0002809791332263242, + "loss": 0.8551, + "step": 180 + }, + { + "epoch": 0.13, + "eval_loss": 0.814437747001648, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 180 + }, + { + "epoch": 0.14, + "eval_loss": 0.8222396969795227, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.317, + "eval_steps_per_second": 6.317, + "step": 190 + }, + { + "epoch": 0.15, + "learning_rate": 0.0002761637239165329, + "loss": 0.8471, + "step": 200 + }, + { + "epoch": 0.15, + "eval_loss": 0.8183712959289551, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 200 + }, + { + "epoch": 0.16, + "eval_loss": 0.8207969665527344, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 210 + }, + { + "epoch": 0.16, + "learning_rate": 0.00027134831460674157, + "loss": 0.8578, + "step": 220 + }, + { + "epoch": 0.16, + "eval_loss": 0.8103113174438477, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 220 + }, + { + "epoch": 0.17, + "eval_loss": 0.808212399482727, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.311, + "eval_steps_per_second": 6.311, + "step": 230 + }, + { + "epoch": 0.18, + "learning_rate": 0.00026653290529695024, + "loss": 0.8389, + "step": 240 + }, + { + "epoch": 0.18, + "eval_loss": 0.7917136549949646, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.351, + "eval_steps_per_second": 6.351, + "step": 240 + }, + { + "epoch": 0.19, + "eval_loss": 0.8259432315826416, + "eval_runtime": 0.1603, + "eval_samples_per_second": 6.24, + "eval_steps_per_second": 6.24, + "step": 250 + }, + { + "epoch": 0.19, + "learning_rate": 0.0002617174959871589, + "loss": 0.8343, + "step": 260 + }, + { + "epoch": 0.19, + "eval_loss": 0.7860772609710693, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.34, + "eval_steps_per_second": 6.34, + "step": 260 + }, + { + "epoch": 0.2, + "eval_loss": 0.8120028972625732, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.347, + "eval_steps_per_second": 6.347, + "step": 270 + }, + { + "epoch": 0.21, + "learning_rate": 0.0002569020866773676, + "loss": 0.831, + "step": 280 + }, + { + "epoch": 0.21, + "eval_loss": 0.7905743718147278, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.368, + "eval_steps_per_second": 6.368, + "step": 280 + }, + { + "epoch": 0.22, + "eval_loss": 0.7759386301040649, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.344, + "eval_steps_per_second": 6.344, + "step": 290 + }, + { + "epoch": 0.22, + "learning_rate": 0.00025208667736757625, + "loss": 0.8325, + "step": 300 + }, + { + "epoch": 0.22, + "eval_loss": 0.8241894245147705, + "eval_runtime": 0.1572, + "eval_samples_per_second": 6.363, + "eval_steps_per_second": 6.363, + "step": 300 + }, + { + "epoch": 0.23, + "eval_loss": 0.7761509418487549, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.346, + "eval_steps_per_second": 6.346, + "step": 310 + }, + { + "epoch": 0.24, + "learning_rate": 0.0002472712680577849, + "loss": 0.8147, + "step": 320 + }, + { + "epoch": 0.24, + "eval_loss": 0.7789342403411865, + "eval_runtime": 0.1572, + "eval_samples_per_second": 6.363, + "eval_steps_per_second": 6.363, + "step": 320 + }, + { + "epoch": 0.25, + "eval_loss": 0.7757179737091064, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.339, + "eval_steps_per_second": 6.339, + "step": 330 + }, + { + "epoch": 0.25, + "learning_rate": 0.00024245585874799357, + "loss": 0.8204, + "step": 340 + }, + { + "epoch": 0.25, + "eval_loss": 0.7954628467559814, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.344, + "eval_steps_per_second": 6.344, + "step": 340 + }, + { + "epoch": 0.26, + "eval_loss": 0.7957539558410645, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 350 + }, + { + "epoch": 0.27, + "learning_rate": 0.00023764044943820224, + "loss": 0.8232, + "step": 360 + }, + { + "epoch": 0.27, + "eval_loss": 0.7863476276397705, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.367, + "eval_steps_per_second": 6.367, + "step": 360 + }, + { + "epoch": 0.27, + "eval_loss": 0.7858812808990479, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.317, + "eval_steps_per_second": 6.317, + "step": 370 + }, + { + "epoch": 0.28, + "learning_rate": 0.00023282504012841088, + "loss": 0.8129, + "step": 380 + }, + { + "epoch": 0.28, + "eval_loss": 0.785178005695343, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.342, + "eval_steps_per_second": 6.342, + "step": 380 + }, + { + "epoch": 0.29, + "eval_loss": 0.8090522289276123, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.318, + "eval_steps_per_second": 6.318, + "step": 390 + }, + { + "epoch": 0.3, + "learning_rate": 0.00022800963081861955, + "loss": 0.8193, + "step": 400 + }, + { + "epoch": 0.3, + "eval_loss": 0.7978605031967163, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.357, + "eval_steps_per_second": 6.357, + "step": 400 + }, + { + "epoch": 0.3, + "eval_loss": 0.7799659371376038, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.327, + "eval_steps_per_second": 6.327, + "step": 410 + }, + { + "epoch": 0.31, + "learning_rate": 0.00022319422150882823, + "loss": 0.8072, + "step": 420 + }, + { + "epoch": 0.31, + "eval_loss": 0.7800132036209106, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.359, + "eval_steps_per_second": 6.359, + "step": 420 + }, + { + "epoch": 0.32, + "eval_loss": 0.7845722436904907, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.329, + "eval_steps_per_second": 6.329, + "step": 430 + }, + { + "epoch": 0.33, + "learning_rate": 0.0002183788121990369, + "loss": 0.8152, + "step": 440 + }, + { + "epoch": 0.33, + "eval_loss": 0.7644073963165283, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.367, + "eval_steps_per_second": 6.367, + "step": 440 + }, + { + "epoch": 0.33, + "eval_loss": 0.7579188346862793, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.367, + "eval_steps_per_second": 6.367, + "step": 450 + }, + { + "epoch": 0.34, + "learning_rate": 0.00021356340288924557, + "loss": 0.8074, + "step": 460 + }, + { + "epoch": 0.34, + "eval_loss": 0.7676059007644653, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.341, + "eval_steps_per_second": 6.341, + "step": 460 + }, + { + "epoch": 0.35, + "eval_loss": 0.7583163976669312, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.358, + "eval_steps_per_second": 6.358, + "step": 470 + }, + { + "epoch": 0.36, + "learning_rate": 0.00020874799357945424, + "loss": 0.8118, + "step": 480 + }, + { + "epoch": 0.36, + "eval_loss": 0.7768086194992065, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.341, + "eval_steps_per_second": 6.341, + "step": 480 + }, + { + "epoch": 0.36, + "eval_loss": 0.7975252270698547, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.37, + "eval_steps_per_second": 6.37, + "step": 490 + }, + { + "epoch": 0.37, + "learning_rate": 0.0002039325842696629, + "loss": 0.8098, + "step": 500 + }, + { + "epoch": 0.37, + "eval_loss": 0.8035451173782349, + "eval_runtime": 0.1565, + "eval_samples_per_second": 6.388, + "eval_steps_per_second": 6.388, + "step": 500 + }, + { + "epoch": 0.38, + "eval_loss": 0.7718653678894043, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.347, + "eval_steps_per_second": 6.347, + "step": 510 + }, + { + "epoch": 0.39, + "learning_rate": 0.00019911717495987158, + "loss": 0.8167, + "step": 520 + }, + { + "epoch": 0.39, + "eval_loss": 0.771480917930603, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.379, + "eval_steps_per_second": 6.379, + "step": 520 + }, + { + "epoch": 0.39, + "eval_loss": 0.7757036685943604, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.364, + "eval_steps_per_second": 6.364, + "step": 530 + }, + { + "epoch": 0.4, + "learning_rate": 0.00019430176565008025, + "loss": 0.8071, + "step": 540 + }, + { + "epoch": 0.4, + "eval_loss": 0.8048112392425537, + "eval_runtime": 0.1563, + "eval_samples_per_second": 6.396, + "eval_steps_per_second": 6.396, + "step": 540 + }, + { + "epoch": 0.41, + "eval_loss": 0.7987676858901978, + "eval_runtime": 0.1562, + "eval_samples_per_second": 6.402, + "eval_steps_per_second": 6.402, + "step": 550 + }, + { + "epoch": 0.42, + "learning_rate": 0.00018948635634028892, + "loss": 0.8143, + "step": 560 + }, + { + "epoch": 0.42, + "eval_loss": 0.7960355281829834, + "eval_runtime": 0.159, + "eval_samples_per_second": 6.29, + "eval_steps_per_second": 6.29, + "step": 560 + }, + { + "epoch": 0.42, + "eval_loss": 0.7673896551132202, + "eval_runtime": 0.1569, + "eval_samples_per_second": 6.372, + "eval_steps_per_second": 6.372, + "step": 570 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001846709470304976, + "loss": 0.7989, + "step": 580 + }, + { + "epoch": 0.43, + "eval_loss": 0.7655194997787476, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.336, + "eval_steps_per_second": 6.336, + "step": 580 + }, + { + "epoch": 0.44, + "eval_loss": 0.7695807218551636, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 6.365, + "step": 590 + }, + { + "epoch": 0.45, + "learning_rate": 0.00017985553772070626, + "loss": 0.8121, + "step": 600 + }, + { + "epoch": 0.45, + "eval_loss": 0.7599303722381592, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.385, + "eval_steps_per_second": 6.385, + "step": 600 + }, + { + "epoch": 0.45, + "eval_loss": 0.7409216165542603, + "eval_runtime": 0.1565, + "eval_samples_per_second": 6.389, + "eval_steps_per_second": 6.389, + "step": 610 + }, + { + "epoch": 0.46, + "learning_rate": 0.00017504012841091494, + "loss": 0.8105, + "step": 620 + }, + { + "epoch": 0.46, + "eval_loss": 0.7620519995689392, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.331, + "eval_steps_per_second": 6.331, + "step": 620 + }, + { + "epoch": 0.47, + "eval_loss": 0.7642089128494263, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.356, + "eval_steps_per_second": 6.356, + "step": 630 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001702247191011236, + "loss": 0.8073, + "step": 640 + }, + { + "epoch": 0.48, + "eval_loss": 0.7464691400527954, + "eval_runtime": 0.156, + "eval_samples_per_second": 6.409, + "eval_steps_per_second": 6.409, + "step": 640 + }, + { + "epoch": 0.48, + "eval_loss": 0.7520545721054077, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.328, + "eval_steps_per_second": 6.328, + "step": 650 + }, + { + "epoch": 0.49, + "learning_rate": 0.00016540930979133222, + "loss": 0.8115, + "step": 660 + }, + { + "epoch": 0.49, + "eval_loss": 0.7851120233535767, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.358, + "eval_steps_per_second": 6.358, + "step": 660 + }, + { + "epoch": 0.5, + "eval_loss": 0.7845426797866821, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.381, + "eval_steps_per_second": 6.381, + "step": 670 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001605939004815409, + "loss": 0.8213, + "step": 680 + }, + { + "epoch": 0.5, + "eval_loss": 0.7726190686225891, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.37, + "eval_steps_per_second": 6.37, + "step": 680 + }, + { + "epoch": 0.51, + "eval_loss": 0.7661857604980469, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.366, + "eval_steps_per_second": 6.366, + "step": 690 + }, + { + "epoch": 0.52, + "learning_rate": 0.00015577849117174957, + "loss": 0.8155, + "step": 700 + }, + { + "epoch": 0.52, + "eval_loss": 0.7846134305000305, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.394, + "eval_steps_per_second": 6.394, + "step": 700 + }, + { + "epoch": 0.53, + "eval_loss": 0.7888688445091248, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.324, + "eval_steps_per_second": 6.324, + "step": 710 + }, + { + "epoch": 0.53, + "learning_rate": 0.00015096308186195824, + "loss": 0.803, + "step": 720 + }, + { + "epoch": 0.53, + "eval_loss": 0.7981730103492737, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.385, + "eval_steps_per_second": 6.385, + "step": 720 + }, + { + "epoch": 0.54, + "eval_loss": 0.8129211664199829, + "eval_runtime": 0.1562, + "eval_samples_per_second": 6.401, + "eval_steps_per_second": 6.401, + "step": 730 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001461476725521669, + "loss": 0.8108, + "step": 740 + }, + { + "epoch": 0.55, + "eval_loss": 0.7717241644859314, + "eval_runtime": 0.1569, + "eval_samples_per_second": 6.375, + "eval_steps_per_second": 6.375, + "step": 740 + }, + { + "epoch": 0.56, + "eval_loss": 0.7536574602127075, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.38, + "eval_steps_per_second": 6.38, + "step": 750 + }, + { + "epoch": 0.56, + "learning_rate": 0.00014133226324237558, + "loss": 0.804, + "step": 760 + }, + { + "epoch": 0.56, + "eval_loss": 0.7452749013900757, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.392, + "eval_steps_per_second": 6.392, + "step": 760 + }, + { + "epoch": 0.57, + "eval_loss": 0.7427541613578796, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.358, + "eval_steps_per_second": 6.358, + "step": 770 + }, + { + "epoch": 0.58, + "learning_rate": 0.00013651685393258425, + "loss": 0.7983, + "step": 780 + }, + { + "epoch": 0.58, + "eval_loss": 0.7469484210014343, + "eval_runtime": 0.1562, + "eval_samples_per_second": 6.404, + "eval_steps_per_second": 6.404, + "step": 780 + }, + { + "epoch": 0.59, + "eval_loss": 0.7667011022567749, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.376, + "eval_steps_per_second": 6.376, + "step": 790 + }, + { + "epoch": 0.59, + "learning_rate": 0.00013170144462279292, + "loss": 0.8054, + "step": 800 + }, + { + "epoch": 0.59, + "eval_loss": 0.7562040090560913, + "eval_runtime": 0.1651, + "eval_samples_per_second": 6.057, + "eval_steps_per_second": 6.057, + "step": 800 + }, + { + "epoch": 0.6, + "eval_loss": 0.7466799020767212, + "eval_runtime": 0.1572, + "eval_samples_per_second": 6.359, + "eval_steps_per_second": 6.359, + "step": 810 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001268860353130016, + "loss": 0.8143, + "step": 820 + }, + { + "epoch": 0.61, + "eval_loss": 0.7601323127746582, + "eval_runtime": 0.1768, + "eval_samples_per_second": 5.657, + "eval_steps_per_second": 5.657, + "step": 820 + }, + { + "epoch": 0.62, + "eval_loss": 0.7430717349052429, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 830 + }, + { + "epoch": 0.62, + "learning_rate": 0.00012207062600321026, + "loss": 0.8038, + "step": 840 + }, + { + "epoch": 0.62, + "eval_loss": 0.7633467316627502, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.342, + "eval_steps_per_second": 6.342, + "step": 840 + }, + { + "epoch": 0.63, + "eval_loss": 0.7451119422912598, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 850 + }, + { + "epoch": 0.64, + "learning_rate": 0.00011725521669341892, + "loss": 0.8053, + "step": 860 + }, + { + "epoch": 0.64, + "eval_loss": 0.7628670930862427, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.338, + "eval_steps_per_second": 6.338, + "step": 860 + }, + { + "epoch": 0.65, + "eval_loss": 0.7490127086639404, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.364, + "eval_steps_per_second": 6.364, + "step": 870 + }, + { + "epoch": 0.65, + "learning_rate": 0.00011243980738362759, + "loss": 0.8007, + "step": 880 + }, + { + "epoch": 0.65, + "eval_loss": 0.767558753490448, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.392, + "eval_steps_per_second": 6.392, + "step": 880 + }, + { + "epoch": 0.66, + "eval_loss": 0.7646104097366333, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 6.365, + "step": 890 + }, + { + "epoch": 0.67, + "learning_rate": 0.00010762439807383626, + "loss": 0.7999, + "step": 900 + }, + { + "epoch": 0.67, + "eval_loss": 0.7580606937408447, + "eval_runtime": 0.1559, + "eval_samples_per_second": 6.414, + "eval_steps_per_second": 6.414, + "step": 900 + }, + { + "epoch": 0.68, + "eval_loss": 0.7789003849029541, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.377, + "eval_steps_per_second": 6.377, + "step": 910 + }, + { + "epoch": 0.68, + "learning_rate": 0.00010280898876404493, + "loss": 0.8007, + "step": 920 + }, + { + "epoch": 0.68, + "eval_loss": 0.7511205077171326, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 6.365, + "step": 920 + }, + { + "epoch": 0.69, + "eval_loss": 0.7637555599212646, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.38, + "eval_steps_per_second": 6.38, + "step": 930 + }, + { + "epoch": 0.7, + "learning_rate": 9.79935794542536e-05, + "loss": 0.8006, + "step": 940 + }, + { + "epoch": 0.7, + "eval_loss": 0.7561322450637817, + "eval_runtime": 0.1582, + "eval_samples_per_second": 6.322, + "eval_steps_per_second": 6.322, + "step": 940 + }, + { + "epoch": 0.71, + "eval_loss": 0.753495454788208, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.395, + "eval_steps_per_second": 6.395, + "step": 950 + }, + { + "epoch": 0.71, + "learning_rate": 9.317817014446228e-05, + "loss": 0.8055, + "step": 960 + }, + { + "epoch": 0.71, + "eval_loss": 0.7372075319290161, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.396, + "eval_steps_per_second": 6.396, + "step": 960 + }, + { + "epoch": 0.72, + "eval_loss": 0.761570394039154, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.329, + "eval_steps_per_second": 6.329, + "step": 970 + }, + { + "epoch": 0.73, + "learning_rate": 8.836276083467093e-05, + "loss": 0.7951, + "step": 980 + }, + { + "epoch": 0.73, + "eval_loss": 0.7430101037025452, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.386, + "eval_steps_per_second": 6.386, + "step": 980 + }, + { + "epoch": 0.74, + "eval_loss": 0.7443385124206543, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.381, + "eval_steps_per_second": 6.381, + "step": 990 + }, + { + "epoch": 0.74, + "learning_rate": 8.35473515248796e-05, + "loss": 0.8049, + "step": 1000 + }, + { + "epoch": 0.74, + "eval_loss": 0.7466243505477905, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.376, + "eval_steps_per_second": 6.376, + "step": 1000 + }, + { + "epoch": 0.75, + "eval_loss": 0.7397581934928894, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.347, + "eval_steps_per_second": 6.347, + "step": 1010 + }, + { + "epoch": 0.76, + "learning_rate": 7.873194221508827e-05, + "loss": 0.8131, + "step": 1020 + }, + { + "epoch": 0.76, + "eval_loss": 0.7271538972854614, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.35, + "eval_steps_per_second": 6.35, + "step": 1020 + }, + { + "epoch": 0.76, + "eval_loss": 0.7354034781455994, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.337, + "eval_steps_per_second": 6.337, + "step": 1030 + }, + { + "epoch": 0.77, + "learning_rate": 7.391653290529695e-05, + "loss": 0.7991, + "step": 1040 + }, + { + "epoch": 0.77, + "eval_loss": 0.7297641038894653, + "eval_runtime": 0.1764, + "eval_samples_per_second": 5.67, + "eval_steps_per_second": 5.67, + "step": 1040 + }, + { + "epoch": 0.78, + "eval_loss": 0.7396876811981201, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.327, + "eval_steps_per_second": 6.327, + "step": 1050 + }, + { + "epoch": 0.79, + "learning_rate": 6.910112359550562e-05, + "loss": 0.8075, + "step": 1060 + }, + { + "epoch": 0.79, + "eval_loss": 0.7296563386917114, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.343, + "eval_steps_per_second": 6.343, + "step": 1060 + }, + { + "epoch": 0.79, + "eval_loss": 0.7264513969421387, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.366, + "eval_steps_per_second": 6.366, + "step": 1070 + } + ], + "max_steps": 1346, + "num_train_epochs": 1, + "total_flos": 1.3908729346916352e+18, + "trial_name": null, + "trial_params": null +} diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1070/training_args.bin b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..2ca60f7ff946be4d5dd08bedeca15733e2f014af --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1070/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b92cf17cb7884e6a8a0c6e72287125f12b4939d9d31cda3054f19a39cab8a0f +size 3579 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1330/optimizer.pt b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..f75aab7964db2191c6c8a35608cf1edb349238c2 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8f90d123b59fb48f1c259db8b47a55a53572ef996f16fec2bedc7fd979af24fd +size 33629893 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1330/pytorch_model.bin b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/pytorch_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..159b78ae74cb23450ff26181c437862856fe5b18 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/pytorch_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c5e38662211e3faf27a44bfabfd0305898802c7ea0a4267fdb810edc645b1abf +size 16822989 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1330/rng_state.pth b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..ff0b006547930bcb275e0505db6cd989d6442d1c --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cfd747162ecefd17bb62c448dfa998491a37c62a7cf84e37e92fe051c6cdd2b0 +size 14575 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1330/scaler.pt b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..74bcb9acd6656c5b10861af13c3552ba64e233b3 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c356c76e72887591ca2fb91ac2075200699081d505caa0a168397f9bf4101a55 +size 557 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1330/scheduler.pt b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..52dfd48373508fd824fbe64d981624750d98ac83 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9b91bd7e6e5acd6fd224a2135a3ad47593def3951c02d2a52653520b1ec933a9 +size 627 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1330/special_tokens_map.json b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..80e8b5de5e5d99c398943f1d5c9a51c3ade8c376 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + } +} diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1330/tokenizer.model b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..6c00c742ce03c627d6cd5b795984876fa49fa899 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347 +size 499723 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1330/tokenizer_config.json b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..381459f4531afd0797d7e2117b21541618a5ad58 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/tokenizer_config.json @@ -0,0 +1,34 @@ +{ + "add_bos_token": true, + "add_eos_token": true, + "bos_token": { + "__type": "AddedToken", + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "clean_up_tokenization_spaces": false, + "eos_token": { + "__type": "AddedToken", + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "model_max_length": 1000000000000000019884624838656, + "pad_token": null, + "sp_model_kwargs": {}, + "special_tokens_map_file": "/home/chiayi/.cache/huggingface/hub/models--decapoda-research--llama-7b-hf/snapshots/5f98eefcc80e437ef68d457ad7bf167c2c6a1348/special_tokens_map.json", + "tokenizer_class": "LlamaTokenizer", + "unk_token": { + "__type": "AddedToken", + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + } +} diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1330/trainer_state.json b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..6cd7dadd33518ed39d6e21353b100d91f23fb67f --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/trainer_state.json @@ -0,0 +1,1476 @@ +{ + "best_metric": 0.7264513969421387, + "best_model_checkpoint": "lora-alpaca-use-seq/checkpoint-1070", + "epoch": 0.9874939093714471, + "global_step": 1330, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.01, + "eval_loss": 2.4221816062927246, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.348, + "eval_steps_per_second": 6.348, + "step": 10 + }, + { + "epoch": 0.01, + "learning_rate": 5.9999999999999995e-05, + "loss": 2.4934, + "step": 20 + }, + { + "epoch": 0.01, + "eval_loss": 2.3228158950805664, + "eval_runtime": 0.1587, + "eval_samples_per_second": 6.302, + "eval_steps_per_second": 6.302, + "step": 20 + }, + { + "epoch": 0.02, + "eval_loss": 1.8552848100662231, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.386, + "eval_steps_per_second": 6.386, + "step": 30 + }, + { + "epoch": 0.03, + "learning_rate": 0.000117, + "loss": 1.9804, + "step": 40 + }, + { + "epoch": 0.03, + "eval_loss": 1.2810053825378418, + "eval_runtime": 0.1559, + "eval_samples_per_second": 6.416, + "eval_steps_per_second": 6.416, + "step": 40 + }, + { + "epoch": 0.04, + "eval_loss": 0.9789324998855591, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.382, + "eval_steps_per_second": 6.382, + "step": 50 + }, + { + "epoch": 0.04, + "learning_rate": 0.00017699999999999997, + "loss": 1.165, + "step": 60 + }, + { + "epoch": 0.04, + "eval_loss": 0.9216629266738892, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.343, + "eval_steps_per_second": 6.343, + "step": 60 + }, + { + "epoch": 0.05, + "eval_loss": 0.8872115612030029, + "eval_runtime": 0.1579, + "eval_samples_per_second": 6.333, + "eval_steps_per_second": 6.333, + "step": 70 + }, + { + "epoch": 0.06, + "learning_rate": 0.000237, + "loss": 0.9931, + "step": 80 + }, + { + "epoch": 0.06, + "eval_loss": 0.8584132194519043, + "eval_runtime": 0.1599, + "eval_samples_per_second": 6.254, + "eval_steps_per_second": 6.254, + "step": 80 + }, + { + "epoch": 0.07, + "eval_loss": 0.9114644527435303, + "eval_runtime": 0.1598, + "eval_samples_per_second": 6.259, + "eval_steps_per_second": 6.259, + "step": 90 + }, + { + "epoch": 0.07, + "learning_rate": 0.00029699999999999996, + "loss": 0.9545, + "step": 100 + }, + { + "epoch": 0.07, + "eval_loss": 0.8909071087837219, + "eval_runtime": 0.1603, + "eval_samples_per_second": 6.24, + "eval_steps_per_second": 6.24, + "step": 100 + }, + { + "epoch": 0.08, + "eval_loss": 0.9121037721633911, + "eval_runtime": 0.1598, + "eval_samples_per_second": 6.259, + "eval_steps_per_second": 6.259, + "step": 110 + }, + { + "epoch": 0.09, + "learning_rate": 0.0002954253611556982, + "loss": 0.8947, + "step": 120 + }, + { + "epoch": 0.09, + "eval_loss": 0.8340358734130859, + "eval_runtime": 0.1605, + "eval_samples_per_second": 6.229, + "eval_steps_per_second": 6.229, + "step": 120 + }, + { + "epoch": 0.1, + "eval_loss": 0.8849254846572876, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.33, + "eval_steps_per_second": 6.33, + "step": 130 + }, + { + "epoch": 0.1, + "learning_rate": 0.0002906099518459069, + "loss": 0.8711, + "step": 140 + }, + { + "epoch": 0.1, + "eval_loss": 0.8834758996963501, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.368, + "eval_steps_per_second": 6.368, + "step": 140 + }, + { + "epoch": 0.11, + "eval_loss": 0.8293902277946472, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.316, + "eval_steps_per_second": 6.316, + "step": 150 + }, + { + "epoch": 0.12, + "learning_rate": 0.00028579454253611556, + "loss": 0.8688, + "step": 160 + }, + { + "epoch": 0.12, + "eval_loss": 0.8478549718856812, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.324, + "eval_steps_per_second": 6.324, + "step": 160 + }, + { + "epoch": 0.13, + "eval_loss": 0.8488869667053223, + "eval_runtime": 0.159, + "eval_samples_per_second": 6.288, + "eval_steps_per_second": 6.288, + "step": 170 + }, + { + "epoch": 0.13, + "learning_rate": 0.0002809791332263242, + "loss": 0.8551, + "step": 180 + }, + { + "epoch": 0.13, + "eval_loss": 0.814437747001648, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 180 + }, + { + "epoch": 0.14, + "eval_loss": 0.8222396969795227, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.317, + "eval_steps_per_second": 6.317, + "step": 190 + }, + { + "epoch": 0.15, + "learning_rate": 0.0002761637239165329, + "loss": 0.8471, + "step": 200 + }, + { + "epoch": 0.15, + "eval_loss": 0.8183712959289551, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 200 + }, + { + "epoch": 0.16, + "eval_loss": 0.8207969665527344, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 210 + }, + { + "epoch": 0.16, + "learning_rate": 0.00027134831460674157, + "loss": 0.8578, + "step": 220 + }, + { + "epoch": 0.16, + "eval_loss": 0.8103113174438477, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 220 + }, + { + "epoch": 0.17, + "eval_loss": 0.808212399482727, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.311, + "eval_steps_per_second": 6.311, + "step": 230 + }, + { + "epoch": 0.18, + "learning_rate": 0.00026653290529695024, + "loss": 0.8389, + "step": 240 + }, + { + "epoch": 0.18, + "eval_loss": 0.7917136549949646, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.351, + "eval_steps_per_second": 6.351, + "step": 240 + }, + { + "epoch": 0.19, + "eval_loss": 0.8259432315826416, + "eval_runtime": 0.1603, + "eval_samples_per_second": 6.24, + "eval_steps_per_second": 6.24, + "step": 250 + }, + { + "epoch": 0.19, + "learning_rate": 0.0002617174959871589, + "loss": 0.8343, + "step": 260 + }, + { + "epoch": 0.19, + "eval_loss": 0.7860772609710693, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.34, + "eval_steps_per_second": 6.34, + "step": 260 + }, + { + "epoch": 0.2, + "eval_loss": 0.8120028972625732, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.347, + "eval_steps_per_second": 6.347, + "step": 270 + }, + { + "epoch": 0.21, + "learning_rate": 0.0002569020866773676, + "loss": 0.831, + "step": 280 + }, + { + "epoch": 0.21, + "eval_loss": 0.7905743718147278, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.368, + "eval_steps_per_second": 6.368, + "step": 280 + }, + { + "epoch": 0.22, + "eval_loss": 0.7759386301040649, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.344, + "eval_steps_per_second": 6.344, + "step": 290 + }, + { + "epoch": 0.22, + "learning_rate": 0.00025208667736757625, + "loss": 0.8325, + "step": 300 + }, + { + "epoch": 0.22, + "eval_loss": 0.8241894245147705, + "eval_runtime": 0.1572, + "eval_samples_per_second": 6.363, + "eval_steps_per_second": 6.363, + "step": 300 + }, + { + "epoch": 0.23, + "eval_loss": 0.7761509418487549, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.346, + "eval_steps_per_second": 6.346, + "step": 310 + }, + { + "epoch": 0.24, + "learning_rate": 0.0002472712680577849, + "loss": 0.8147, + "step": 320 + }, + { + "epoch": 0.24, + "eval_loss": 0.7789342403411865, + "eval_runtime": 0.1572, + "eval_samples_per_second": 6.363, + "eval_steps_per_second": 6.363, + "step": 320 + }, + { + "epoch": 0.25, + "eval_loss": 0.7757179737091064, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.339, + "eval_steps_per_second": 6.339, + "step": 330 + }, + { + "epoch": 0.25, + "learning_rate": 0.00024245585874799357, + "loss": 0.8204, + "step": 340 + }, + { + "epoch": 0.25, + "eval_loss": 0.7954628467559814, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.344, + "eval_steps_per_second": 6.344, + "step": 340 + }, + { + "epoch": 0.26, + "eval_loss": 0.7957539558410645, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 350 + }, + { + "epoch": 0.27, + "learning_rate": 0.00023764044943820224, + "loss": 0.8232, + "step": 360 + }, + { + "epoch": 0.27, + "eval_loss": 0.7863476276397705, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.367, + "eval_steps_per_second": 6.367, + "step": 360 + }, + { + "epoch": 0.27, + "eval_loss": 0.7858812808990479, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.317, + "eval_steps_per_second": 6.317, + "step": 370 + }, + { + "epoch": 0.28, + "learning_rate": 0.00023282504012841088, + "loss": 0.8129, + "step": 380 + }, + { + "epoch": 0.28, + "eval_loss": 0.785178005695343, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.342, + "eval_steps_per_second": 6.342, + "step": 380 + }, + { + "epoch": 0.29, + "eval_loss": 0.8090522289276123, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.318, + "eval_steps_per_second": 6.318, + "step": 390 + }, + { + "epoch": 0.3, + "learning_rate": 0.00022800963081861955, + "loss": 0.8193, + "step": 400 + }, + { + "epoch": 0.3, + "eval_loss": 0.7978605031967163, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.357, + "eval_steps_per_second": 6.357, + "step": 400 + }, + { + "epoch": 0.3, + "eval_loss": 0.7799659371376038, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.327, + "eval_steps_per_second": 6.327, + "step": 410 + }, + { + "epoch": 0.31, + "learning_rate": 0.00022319422150882823, + "loss": 0.8072, + "step": 420 + }, + { + "epoch": 0.31, + "eval_loss": 0.7800132036209106, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.359, + "eval_steps_per_second": 6.359, + "step": 420 + }, + { + "epoch": 0.32, + "eval_loss": 0.7845722436904907, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.329, + "eval_steps_per_second": 6.329, + "step": 430 + }, + { + "epoch": 0.33, + "learning_rate": 0.0002183788121990369, + "loss": 0.8152, + "step": 440 + }, + { + "epoch": 0.33, + "eval_loss": 0.7644073963165283, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.367, + "eval_steps_per_second": 6.367, + "step": 440 + }, + { + "epoch": 0.33, + "eval_loss": 0.7579188346862793, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.367, + "eval_steps_per_second": 6.367, + "step": 450 + }, + { + "epoch": 0.34, + "learning_rate": 0.00021356340288924557, + "loss": 0.8074, + "step": 460 + }, + { + "epoch": 0.34, + "eval_loss": 0.7676059007644653, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.341, + "eval_steps_per_second": 6.341, + "step": 460 + }, + { + "epoch": 0.35, + "eval_loss": 0.7583163976669312, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.358, + "eval_steps_per_second": 6.358, + "step": 470 + }, + { + "epoch": 0.36, + "learning_rate": 0.00020874799357945424, + "loss": 0.8118, + "step": 480 + }, + { + "epoch": 0.36, + "eval_loss": 0.7768086194992065, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.341, + "eval_steps_per_second": 6.341, + "step": 480 + }, + { + "epoch": 0.36, + "eval_loss": 0.7975252270698547, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.37, + "eval_steps_per_second": 6.37, + "step": 490 + }, + { + "epoch": 0.37, + "learning_rate": 0.0002039325842696629, + "loss": 0.8098, + "step": 500 + }, + { + "epoch": 0.37, + "eval_loss": 0.8035451173782349, + "eval_runtime": 0.1565, + "eval_samples_per_second": 6.388, + "eval_steps_per_second": 6.388, + "step": 500 + }, + { + "epoch": 0.38, + "eval_loss": 0.7718653678894043, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.347, + "eval_steps_per_second": 6.347, + "step": 510 + }, + { + "epoch": 0.39, + "learning_rate": 0.00019911717495987158, + "loss": 0.8167, + "step": 520 + }, + { + "epoch": 0.39, + "eval_loss": 0.771480917930603, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.379, + "eval_steps_per_second": 6.379, + "step": 520 + }, + { + "epoch": 0.39, + "eval_loss": 0.7757036685943604, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.364, + "eval_steps_per_second": 6.364, + "step": 530 + }, + { + "epoch": 0.4, + "learning_rate": 0.00019430176565008025, + "loss": 0.8071, + "step": 540 + }, + { + "epoch": 0.4, + "eval_loss": 0.8048112392425537, + "eval_runtime": 0.1563, + "eval_samples_per_second": 6.396, + "eval_steps_per_second": 6.396, + "step": 540 + }, + { + "epoch": 0.41, + "eval_loss": 0.7987676858901978, + "eval_runtime": 0.1562, + "eval_samples_per_second": 6.402, + "eval_steps_per_second": 6.402, + "step": 550 + }, + { + "epoch": 0.42, + "learning_rate": 0.00018948635634028892, + "loss": 0.8143, + "step": 560 + }, + { + "epoch": 0.42, + "eval_loss": 0.7960355281829834, + "eval_runtime": 0.159, + "eval_samples_per_second": 6.29, + "eval_steps_per_second": 6.29, + "step": 560 + }, + { + "epoch": 0.42, + "eval_loss": 0.7673896551132202, + "eval_runtime": 0.1569, + "eval_samples_per_second": 6.372, + "eval_steps_per_second": 6.372, + "step": 570 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001846709470304976, + "loss": 0.7989, + "step": 580 + }, + { + "epoch": 0.43, + "eval_loss": 0.7655194997787476, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.336, + "eval_steps_per_second": 6.336, + "step": 580 + }, + { + "epoch": 0.44, + "eval_loss": 0.7695807218551636, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 6.365, + "step": 590 + }, + { + "epoch": 0.45, + "learning_rate": 0.00017985553772070626, + "loss": 0.8121, + "step": 600 + }, + { + "epoch": 0.45, + "eval_loss": 0.7599303722381592, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.385, + "eval_steps_per_second": 6.385, + "step": 600 + }, + { + "epoch": 0.45, + "eval_loss": 0.7409216165542603, + "eval_runtime": 0.1565, + "eval_samples_per_second": 6.389, + "eval_steps_per_second": 6.389, + "step": 610 + }, + { + "epoch": 0.46, + "learning_rate": 0.00017504012841091494, + "loss": 0.8105, + "step": 620 + }, + { + "epoch": 0.46, + "eval_loss": 0.7620519995689392, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.331, + "eval_steps_per_second": 6.331, + "step": 620 + }, + { + "epoch": 0.47, + "eval_loss": 0.7642089128494263, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.356, + "eval_steps_per_second": 6.356, + "step": 630 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001702247191011236, + "loss": 0.8073, + "step": 640 + }, + { + "epoch": 0.48, + "eval_loss": 0.7464691400527954, + "eval_runtime": 0.156, + "eval_samples_per_second": 6.409, + "eval_steps_per_second": 6.409, + "step": 640 + }, + { + "epoch": 0.48, + "eval_loss": 0.7520545721054077, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.328, + "eval_steps_per_second": 6.328, + "step": 650 + }, + { + "epoch": 0.49, + "learning_rate": 0.00016540930979133222, + "loss": 0.8115, + "step": 660 + }, + { + "epoch": 0.49, + "eval_loss": 0.7851120233535767, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.358, + "eval_steps_per_second": 6.358, + "step": 660 + }, + { + "epoch": 0.5, + "eval_loss": 0.7845426797866821, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.381, + "eval_steps_per_second": 6.381, + "step": 670 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001605939004815409, + "loss": 0.8213, + "step": 680 + }, + { + "epoch": 0.5, + "eval_loss": 0.7726190686225891, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.37, + "eval_steps_per_second": 6.37, + "step": 680 + }, + { + "epoch": 0.51, + "eval_loss": 0.7661857604980469, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.366, + "eval_steps_per_second": 6.366, + "step": 690 + }, + { + "epoch": 0.52, + "learning_rate": 0.00015577849117174957, + "loss": 0.8155, + "step": 700 + }, + { + "epoch": 0.52, + "eval_loss": 0.7846134305000305, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.394, + "eval_steps_per_second": 6.394, + "step": 700 + }, + { + "epoch": 0.53, + "eval_loss": 0.7888688445091248, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.324, + "eval_steps_per_second": 6.324, + "step": 710 + }, + { + "epoch": 0.53, + "learning_rate": 0.00015096308186195824, + "loss": 0.803, + "step": 720 + }, + { + "epoch": 0.53, + "eval_loss": 0.7981730103492737, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.385, + "eval_steps_per_second": 6.385, + "step": 720 + }, + { + "epoch": 0.54, + "eval_loss": 0.8129211664199829, + "eval_runtime": 0.1562, + "eval_samples_per_second": 6.401, + "eval_steps_per_second": 6.401, + "step": 730 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001461476725521669, + "loss": 0.8108, + "step": 740 + }, + { + "epoch": 0.55, + "eval_loss": 0.7717241644859314, + "eval_runtime": 0.1569, + "eval_samples_per_second": 6.375, + "eval_steps_per_second": 6.375, + "step": 740 + }, + { + "epoch": 0.56, + "eval_loss": 0.7536574602127075, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.38, + "eval_steps_per_second": 6.38, + "step": 750 + }, + { + "epoch": 0.56, + "learning_rate": 0.00014133226324237558, + "loss": 0.804, + "step": 760 + }, + { + "epoch": 0.56, + "eval_loss": 0.7452749013900757, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.392, + "eval_steps_per_second": 6.392, + "step": 760 + }, + { + "epoch": 0.57, + "eval_loss": 0.7427541613578796, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.358, + "eval_steps_per_second": 6.358, + "step": 770 + }, + { + "epoch": 0.58, + "learning_rate": 0.00013651685393258425, + "loss": 0.7983, + "step": 780 + }, + { + "epoch": 0.58, + "eval_loss": 0.7469484210014343, + "eval_runtime": 0.1562, + "eval_samples_per_second": 6.404, + "eval_steps_per_second": 6.404, + "step": 780 + }, + { + "epoch": 0.59, + "eval_loss": 0.7667011022567749, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.376, + "eval_steps_per_second": 6.376, + "step": 790 + }, + { + "epoch": 0.59, + "learning_rate": 0.00013170144462279292, + "loss": 0.8054, + "step": 800 + }, + { + "epoch": 0.59, + "eval_loss": 0.7562040090560913, + "eval_runtime": 0.1651, + "eval_samples_per_second": 6.057, + "eval_steps_per_second": 6.057, + "step": 800 + }, + { + "epoch": 0.6, + "eval_loss": 0.7466799020767212, + "eval_runtime": 0.1572, + "eval_samples_per_second": 6.359, + "eval_steps_per_second": 6.359, + "step": 810 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001268860353130016, + "loss": 0.8143, + "step": 820 + }, + { + "epoch": 0.61, + "eval_loss": 0.7601323127746582, + "eval_runtime": 0.1768, + "eval_samples_per_second": 5.657, + "eval_steps_per_second": 5.657, + "step": 820 + }, + { + "epoch": 0.62, + "eval_loss": 0.7430717349052429, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 830 + }, + { + "epoch": 0.62, + "learning_rate": 0.00012207062600321026, + "loss": 0.8038, + "step": 840 + }, + { + "epoch": 0.62, + "eval_loss": 0.7633467316627502, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.342, + "eval_steps_per_second": 6.342, + "step": 840 + }, + { + "epoch": 0.63, + "eval_loss": 0.7451119422912598, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 850 + }, + { + "epoch": 0.64, + "learning_rate": 0.00011725521669341892, + "loss": 0.8053, + "step": 860 + }, + { + "epoch": 0.64, + "eval_loss": 0.7628670930862427, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.338, + "eval_steps_per_second": 6.338, + "step": 860 + }, + { + "epoch": 0.65, + "eval_loss": 0.7490127086639404, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.364, + "eval_steps_per_second": 6.364, + "step": 870 + }, + { + "epoch": 0.65, + "learning_rate": 0.00011243980738362759, + "loss": 0.8007, + "step": 880 + }, + { + "epoch": 0.65, + "eval_loss": 0.767558753490448, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.392, + "eval_steps_per_second": 6.392, + "step": 880 + }, + { + "epoch": 0.66, + "eval_loss": 0.7646104097366333, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 6.365, + "step": 890 + }, + { + "epoch": 0.67, + "learning_rate": 0.00010762439807383626, + "loss": 0.7999, + "step": 900 + }, + { + "epoch": 0.67, + "eval_loss": 0.7580606937408447, + "eval_runtime": 0.1559, + "eval_samples_per_second": 6.414, + "eval_steps_per_second": 6.414, + "step": 900 + }, + { + "epoch": 0.68, + "eval_loss": 0.7789003849029541, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.377, + "eval_steps_per_second": 6.377, + "step": 910 + }, + { + "epoch": 0.68, + "learning_rate": 0.00010280898876404493, + "loss": 0.8007, + "step": 920 + }, + { + "epoch": 0.68, + "eval_loss": 0.7511205077171326, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 6.365, + "step": 920 + }, + { + "epoch": 0.69, + "eval_loss": 0.7637555599212646, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.38, + "eval_steps_per_second": 6.38, + "step": 930 + }, + { + "epoch": 0.7, + "learning_rate": 9.79935794542536e-05, + "loss": 0.8006, + "step": 940 + }, + { + "epoch": 0.7, + "eval_loss": 0.7561322450637817, + "eval_runtime": 0.1582, + "eval_samples_per_second": 6.322, + "eval_steps_per_second": 6.322, + "step": 940 + }, + { + "epoch": 0.71, + "eval_loss": 0.753495454788208, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.395, + "eval_steps_per_second": 6.395, + "step": 950 + }, + { + "epoch": 0.71, + "learning_rate": 9.317817014446228e-05, + "loss": 0.8055, + "step": 960 + }, + { + "epoch": 0.71, + "eval_loss": 0.7372075319290161, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.396, + "eval_steps_per_second": 6.396, + "step": 960 + }, + { + "epoch": 0.72, + "eval_loss": 0.761570394039154, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.329, + "eval_steps_per_second": 6.329, + "step": 970 + }, + { + "epoch": 0.73, + "learning_rate": 8.836276083467093e-05, + "loss": 0.7951, + "step": 980 + }, + { + "epoch": 0.73, + "eval_loss": 0.7430101037025452, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.386, + "eval_steps_per_second": 6.386, + "step": 980 + }, + { + "epoch": 0.74, + "eval_loss": 0.7443385124206543, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.381, + "eval_steps_per_second": 6.381, + "step": 990 + }, + { + "epoch": 0.74, + "learning_rate": 8.35473515248796e-05, + "loss": 0.8049, + "step": 1000 + }, + { + "epoch": 0.74, + "eval_loss": 0.7466243505477905, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.376, + "eval_steps_per_second": 6.376, + "step": 1000 + }, + { + "epoch": 0.75, + "eval_loss": 0.7397581934928894, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.347, + "eval_steps_per_second": 6.347, + "step": 1010 + }, + { + "epoch": 0.76, + "learning_rate": 7.873194221508827e-05, + "loss": 0.8131, + "step": 1020 + }, + { + "epoch": 0.76, + "eval_loss": 0.7271538972854614, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.35, + "eval_steps_per_second": 6.35, + "step": 1020 + }, + { + "epoch": 0.76, + "eval_loss": 0.7354034781455994, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.337, + "eval_steps_per_second": 6.337, + "step": 1030 + }, + { + "epoch": 0.77, + "learning_rate": 7.391653290529695e-05, + "loss": 0.7991, + "step": 1040 + }, + { + "epoch": 0.77, + "eval_loss": 0.7297641038894653, + "eval_runtime": 0.1764, + "eval_samples_per_second": 5.67, + "eval_steps_per_second": 5.67, + "step": 1040 + }, + { + "epoch": 0.78, + "eval_loss": 0.7396876811981201, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.327, + "eval_steps_per_second": 6.327, + "step": 1050 + }, + { + "epoch": 0.79, + "learning_rate": 6.910112359550562e-05, + "loss": 0.8075, + "step": 1060 + }, + { + "epoch": 0.79, + "eval_loss": 0.7296563386917114, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.343, + "eval_steps_per_second": 6.343, + "step": 1060 + }, + { + "epoch": 0.79, + "eval_loss": 0.7264513969421387, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.366, + "eval_steps_per_second": 6.366, + "step": 1070 + }, + { + "epoch": 0.8, + "learning_rate": 6.428571428571427e-05, + "loss": 0.8075, + "step": 1080 + }, + { + "epoch": 0.8, + "eval_loss": 0.7289776802062988, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.367, + "eval_steps_per_second": 6.367, + "step": 1080 + }, + { + "epoch": 0.81, + "eval_loss": 0.7540558576583862, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 6.365, + "step": 1090 + }, + { + "epoch": 0.82, + "learning_rate": 5.9470304975922945e-05, + "loss": 0.7916, + "step": 1100 + }, + { + "epoch": 0.82, + "eval_loss": 0.7357069253921509, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.393, + "eval_steps_per_second": 6.393, + "step": 1100 + }, + { + "epoch": 0.82, + "eval_loss": 0.7496874928474426, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.351, + "eval_steps_per_second": 6.351, + "step": 1110 + }, + { + "epoch": 0.83, + "learning_rate": 5.4654895666131616e-05, + "loss": 0.7899, + "step": 1120 + }, + { + "epoch": 0.83, + "eval_loss": 0.7296478748321533, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.357, + "eval_steps_per_second": 6.357, + "step": 1120 + }, + { + "epoch": 0.84, + "eval_loss": 0.7285305261611938, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.346, + "eval_steps_per_second": 6.346, + "step": 1130 + }, + { + "epoch": 0.85, + "learning_rate": 4.983948635634029e-05, + "loss": 0.7958, + "step": 1140 + }, + { + "epoch": 0.85, + "eval_loss": 0.7548575401306152, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.368, + "eval_steps_per_second": 6.368, + "step": 1140 + }, + { + "epoch": 0.85, + "eval_loss": 0.75063556432724, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.343, + "eval_steps_per_second": 6.343, + "step": 1150 + }, + { + "epoch": 0.86, + "learning_rate": 4.502407704654895e-05, + "loss": 0.8004, + "step": 1160 + }, + { + "epoch": 0.86, + "eval_loss": 0.7478867769241333, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.341, + "eval_steps_per_second": 6.341, + "step": 1160 + }, + { + "epoch": 0.87, + "eval_loss": 0.7417320609092712, + "eval_runtime": 0.1572, + "eval_samples_per_second": 6.361, + "eval_steps_per_second": 6.361, + "step": 1170 + }, + { + "epoch": 0.88, + "learning_rate": 4.020866773675762e-05, + "loss": 0.7927, + "step": 1180 + }, + { + "epoch": 0.88, + "eval_loss": 0.7448097467422485, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.337, + "eval_steps_per_second": 6.337, + "step": 1180 + }, + { + "epoch": 0.88, + "eval_loss": 0.7570788860321045, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.346, + "eval_steps_per_second": 6.346, + "step": 1190 + }, + { + "epoch": 0.89, + "learning_rate": 3.539325842696629e-05, + "loss": 0.8013, + "step": 1200 + }, + { + "epoch": 0.89, + "eval_loss": 0.7594838738441467, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.324, + "eval_steps_per_second": 6.324, + "step": 1200 + }, + { + "epoch": 0.9, + "eval_loss": 0.7382398843765259, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.369, + "eval_steps_per_second": 6.369, + "step": 1210 + }, + { + "epoch": 0.91, + "learning_rate": 3.057784911717496e-05, + "loss": 0.8002, + "step": 1220 + }, + { + "epoch": 0.91, + "eval_loss": 0.7314504384994507, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.366, + "eval_steps_per_second": 6.366, + "step": 1220 + }, + { + "epoch": 0.91, + "eval_loss": 0.780935525894165, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.348, + "eval_steps_per_second": 6.348, + "step": 1230 + }, + { + "epoch": 0.92, + "learning_rate": 2.5762439807383626e-05, + "loss": 0.7914, + "step": 1240 + }, + { + "epoch": 0.92, + "eval_loss": 0.7425116300582886, + "eval_runtime": 0.1574, + "eval_samples_per_second": 6.352, + "eval_steps_per_second": 6.352, + "step": 1240 + }, + { + "epoch": 0.93, + "eval_loss": 0.7363260984420776, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.394, + "eval_steps_per_second": 6.394, + "step": 1250 + }, + { + "epoch": 0.94, + "learning_rate": 2.0947030497592293e-05, + "loss": 0.7911, + "step": 1260 + }, + { + "epoch": 0.94, + "eval_loss": 0.7363616228103638, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.385, + "eval_steps_per_second": 6.385, + "step": 1260 + }, + { + "epoch": 0.94, + "eval_loss": 0.7710204124450684, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 1270 + }, + { + "epoch": 0.95, + "learning_rate": 1.613162118780096e-05, + "loss": 0.7973, + "step": 1280 + }, + { + "epoch": 0.95, + "eval_loss": 0.7362488508224487, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 1280 + }, + { + "epoch": 0.96, + "eval_loss": 0.7484357357025146, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.326, + "eval_steps_per_second": 6.326, + "step": 1290 + }, + { + "epoch": 0.97, + "learning_rate": 1.1316211878009629e-05, + "loss": 0.8008, + "step": 1300 + }, + { + "epoch": 0.97, + "eval_loss": 0.7393674254417419, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.338, + "eval_steps_per_second": 6.338, + "step": 1300 + }, + { + "epoch": 0.97, + "eval_loss": 0.738249659538269, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.344, + "eval_steps_per_second": 6.344, + "step": 1310 + }, + { + "epoch": 0.98, + "learning_rate": 6.500802568218298e-06, + "loss": 0.798, + "step": 1320 + }, + { + "epoch": 0.98, + "eval_loss": 0.7329107522964478, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.351, + "eval_steps_per_second": 6.351, + "step": 1320 + }, + { + "epoch": 0.99, + "eval_loss": 0.7391780614852905, + "eval_runtime": 0.1587, + "eval_samples_per_second": 6.302, + "eval_steps_per_second": 6.302, + "step": 1330 + } + ], + "max_steps": 1346, + "num_train_epochs": 1, + "total_flos": 1.7288420590092288e+18, + "trial_name": null, + "trial_params": null +} diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1330/training_args.bin b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..2ca60f7ff946be4d5dd08bedeca15733e2f014af --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1330/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b92cf17cb7884e6a8a0c6e72287125f12b4939d9d31cda3054f19a39cab8a0f +size 3579 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1340/optimizer.pt b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..8fbc1104aa76085d2880dd0a6dfb5c5be617ca1f --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5cf06a5c2e8156cc2c7229db25719db040a0ac9e10addbb17057eabeedd96070 +size 33629893 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1340/pytorch_model.bin b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/pytorch_model.bin new file mode 100644 index 0000000000000000000000000000000000000000..b073e30d5e8d218d281067ec138d38888d976716 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/pytorch_model.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:82f64e634d406a6c87c3a7f9b5a50fe156dce69f19f1b3f96bfb68cf4b2ca3e1 +size 16822989 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1340/rng_state.pth b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/rng_state.pth new file mode 100644 index 0000000000000000000000000000000000000000..33da5750e31afbc0dc4b497d5e0535693c5264af --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/rng_state.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:19eebdc2da227039d15d23eda2cacede2cbb39408ff2226e4b64264dd8f643b9 +size 14575 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1340/scaler.pt b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/scaler.pt new file mode 100644 index 0000000000000000000000000000000000000000..336c82e86d43ca8802491f53e4ae650a51081ec3 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/scaler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c7df3b95085abe7783222569cc0c4d99aa942ba8993fe5b40b3781470aea56cf +size 557 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1340/scheduler.pt b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..698288887a058413b4142084c3e030960afa3767 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c1ac63ec028cab7e73283576b684f7acdeef93912e5f833ba54976988f175fb4 +size 627 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1340/special_tokens_map.json b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..80e8b5de5e5d99c398943f1d5c9a51c3ade8c376 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + } +} diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1340/tokenizer.model b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..6c00c742ce03c627d6cd5b795984876fa49fa899 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347 +size 499723 diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1340/tokenizer_config.json b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..381459f4531afd0797d7e2117b21541618a5ad58 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/tokenizer_config.json @@ -0,0 +1,34 @@ +{ + "add_bos_token": true, + "add_eos_token": true, + "bos_token": { + "__type": "AddedToken", + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "clean_up_tokenization_spaces": false, + "eos_token": { + "__type": "AddedToken", + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + }, + "model_max_length": 1000000000000000019884624838656, + "pad_token": null, + "sp_model_kwargs": {}, + "special_tokens_map_file": "/home/chiayi/.cache/huggingface/hub/models--decapoda-research--llama-7b-hf/snapshots/5f98eefcc80e437ef68d457ad7bf167c2c6a1348/special_tokens_map.json", + "tokenizer_class": "LlamaTokenizer", + "unk_token": { + "__type": "AddedToken", + "content": "", + "lstrip": false, + "normalized": true, + "rstrip": false, + "single_word": false + } +} diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1340/trainer_state.json b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..49e659078eaa88f1240da638e3f933c5f53dcffe --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/trainer_state.json @@ -0,0 +1,1490 @@ +{ + "best_metric": 0.7264513969421387, + "best_model_checkpoint": "lora-alpaca-use-seq/checkpoint-1070", + "epoch": 0.9949186756073227, + "global_step": 1340, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.01, + "eval_loss": 2.4221816062927246, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.348, + "eval_steps_per_second": 6.348, + "step": 10 + }, + { + "epoch": 0.01, + "learning_rate": 5.9999999999999995e-05, + "loss": 2.4934, + "step": 20 + }, + { + "epoch": 0.01, + "eval_loss": 2.3228158950805664, + "eval_runtime": 0.1587, + "eval_samples_per_second": 6.302, + "eval_steps_per_second": 6.302, + "step": 20 + }, + { + "epoch": 0.02, + "eval_loss": 1.8552848100662231, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.386, + "eval_steps_per_second": 6.386, + "step": 30 + }, + { + "epoch": 0.03, + "learning_rate": 0.000117, + "loss": 1.9804, + "step": 40 + }, + { + "epoch": 0.03, + "eval_loss": 1.2810053825378418, + "eval_runtime": 0.1559, + "eval_samples_per_second": 6.416, + "eval_steps_per_second": 6.416, + "step": 40 + }, + { + "epoch": 0.04, + "eval_loss": 0.9789324998855591, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.382, + "eval_steps_per_second": 6.382, + "step": 50 + }, + { + "epoch": 0.04, + "learning_rate": 0.00017699999999999997, + "loss": 1.165, + "step": 60 + }, + { + "epoch": 0.04, + "eval_loss": 0.9216629266738892, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.343, + "eval_steps_per_second": 6.343, + "step": 60 + }, + { + "epoch": 0.05, + "eval_loss": 0.8872115612030029, + "eval_runtime": 0.1579, + "eval_samples_per_second": 6.333, + "eval_steps_per_second": 6.333, + "step": 70 + }, + { + "epoch": 0.06, + "learning_rate": 0.000237, + "loss": 0.9931, + "step": 80 + }, + { + "epoch": 0.06, + "eval_loss": 0.8584132194519043, + "eval_runtime": 0.1599, + "eval_samples_per_second": 6.254, + "eval_steps_per_second": 6.254, + "step": 80 + }, + { + "epoch": 0.07, + "eval_loss": 0.9114644527435303, + "eval_runtime": 0.1598, + "eval_samples_per_second": 6.259, + "eval_steps_per_second": 6.259, + "step": 90 + }, + { + "epoch": 0.07, + "learning_rate": 0.00029699999999999996, + "loss": 0.9545, + "step": 100 + }, + { + "epoch": 0.07, + "eval_loss": 0.8909071087837219, + "eval_runtime": 0.1603, + "eval_samples_per_second": 6.24, + "eval_steps_per_second": 6.24, + "step": 100 + }, + { + "epoch": 0.08, + "eval_loss": 0.9121037721633911, + "eval_runtime": 0.1598, + "eval_samples_per_second": 6.259, + "eval_steps_per_second": 6.259, + "step": 110 + }, + { + "epoch": 0.09, + "learning_rate": 0.0002954253611556982, + "loss": 0.8947, + "step": 120 + }, + { + "epoch": 0.09, + "eval_loss": 0.8340358734130859, + "eval_runtime": 0.1605, + "eval_samples_per_second": 6.229, + "eval_steps_per_second": 6.229, + "step": 120 + }, + { + "epoch": 0.1, + "eval_loss": 0.8849254846572876, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.33, + "eval_steps_per_second": 6.33, + "step": 130 + }, + { + "epoch": 0.1, + "learning_rate": 0.0002906099518459069, + "loss": 0.8711, + "step": 140 + }, + { + "epoch": 0.1, + "eval_loss": 0.8834758996963501, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.368, + "eval_steps_per_second": 6.368, + "step": 140 + }, + { + "epoch": 0.11, + "eval_loss": 0.8293902277946472, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.316, + "eval_steps_per_second": 6.316, + "step": 150 + }, + { + "epoch": 0.12, + "learning_rate": 0.00028579454253611556, + "loss": 0.8688, + "step": 160 + }, + { + "epoch": 0.12, + "eval_loss": 0.8478549718856812, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.324, + "eval_steps_per_second": 6.324, + "step": 160 + }, + { + "epoch": 0.13, + "eval_loss": 0.8488869667053223, + "eval_runtime": 0.159, + "eval_samples_per_second": 6.288, + "eval_steps_per_second": 6.288, + "step": 170 + }, + { + "epoch": 0.13, + "learning_rate": 0.0002809791332263242, + "loss": 0.8551, + "step": 180 + }, + { + "epoch": 0.13, + "eval_loss": 0.814437747001648, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 180 + }, + { + "epoch": 0.14, + "eval_loss": 0.8222396969795227, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.317, + "eval_steps_per_second": 6.317, + "step": 190 + }, + { + "epoch": 0.15, + "learning_rate": 0.0002761637239165329, + "loss": 0.8471, + "step": 200 + }, + { + "epoch": 0.15, + "eval_loss": 0.8183712959289551, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 200 + }, + { + "epoch": 0.16, + "eval_loss": 0.8207969665527344, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 210 + }, + { + "epoch": 0.16, + "learning_rate": 0.00027134831460674157, + "loss": 0.8578, + "step": 220 + }, + { + "epoch": 0.16, + "eval_loss": 0.8103113174438477, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 220 + }, + { + "epoch": 0.17, + "eval_loss": 0.808212399482727, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.311, + "eval_steps_per_second": 6.311, + "step": 230 + }, + { + "epoch": 0.18, + "learning_rate": 0.00026653290529695024, + "loss": 0.8389, + "step": 240 + }, + { + "epoch": 0.18, + "eval_loss": 0.7917136549949646, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.351, + "eval_steps_per_second": 6.351, + "step": 240 + }, + { + "epoch": 0.19, + "eval_loss": 0.8259432315826416, + "eval_runtime": 0.1603, + "eval_samples_per_second": 6.24, + "eval_steps_per_second": 6.24, + "step": 250 + }, + { + "epoch": 0.19, + "learning_rate": 0.0002617174959871589, + "loss": 0.8343, + "step": 260 + }, + { + "epoch": 0.19, + "eval_loss": 0.7860772609710693, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.34, + "eval_steps_per_second": 6.34, + "step": 260 + }, + { + "epoch": 0.2, + "eval_loss": 0.8120028972625732, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.347, + "eval_steps_per_second": 6.347, + "step": 270 + }, + { + "epoch": 0.21, + "learning_rate": 0.0002569020866773676, + "loss": 0.831, + "step": 280 + }, + { + "epoch": 0.21, + "eval_loss": 0.7905743718147278, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.368, + "eval_steps_per_second": 6.368, + "step": 280 + }, + { + "epoch": 0.22, + "eval_loss": 0.7759386301040649, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.344, + "eval_steps_per_second": 6.344, + "step": 290 + }, + { + "epoch": 0.22, + "learning_rate": 0.00025208667736757625, + "loss": 0.8325, + "step": 300 + }, + { + "epoch": 0.22, + "eval_loss": 0.8241894245147705, + "eval_runtime": 0.1572, + "eval_samples_per_second": 6.363, + "eval_steps_per_second": 6.363, + "step": 300 + }, + { + "epoch": 0.23, + "eval_loss": 0.7761509418487549, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.346, + "eval_steps_per_second": 6.346, + "step": 310 + }, + { + "epoch": 0.24, + "learning_rate": 0.0002472712680577849, + "loss": 0.8147, + "step": 320 + }, + { + "epoch": 0.24, + "eval_loss": 0.7789342403411865, + "eval_runtime": 0.1572, + "eval_samples_per_second": 6.363, + "eval_steps_per_second": 6.363, + "step": 320 + }, + { + "epoch": 0.25, + "eval_loss": 0.7757179737091064, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.339, + "eval_steps_per_second": 6.339, + "step": 330 + }, + { + "epoch": 0.25, + "learning_rate": 0.00024245585874799357, + "loss": 0.8204, + "step": 340 + }, + { + "epoch": 0.25, + "eval_loss": 0.7954628467559814, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.344, + "eval_steps_per_second": 6.344, + "step": 340 + }, + { + "epoch": 0.26, + "eval_loss": 0.7957539558410645, + "eval_runtime": 0.1584, + "eval_samples_per_second": 6.315, + "eval_steps_per_second": 6.315, + "step": 350 + }, + { + "epoch": 0.27, + "learning_rate": 0.00023764044943820224, + "loss": 0.8232, + "step": 360 + }, + { + "epoch": 0.27, + "eval_loss": 0.7863476276397705, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.367, + "eval_steps_per_second": 6.367, + "step": 360 + }, + { + "epoch": 0.27, + "eval_loss": 0.7858812808990479, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.317, + "eval_steps_per_second": 6.317, + "step": 370 + }, + { + "epoch": 0.28, + "learning_rate": 0.00023282504012841088, + "loss": 0.8129, + "step": 380 + }, + { + "epoch": 0.28, + "eval_loss": 0.785178005695343, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.342, + "eval_steps_per_second": 6.342, + "step": 380 + }, + { + "epoch": 0.29, + "eval_loss": 0.8090522289276123, + "eval_runtime": 0.1583, + "eval_samples_per_second": 6.318, + "eval_steps_per_second": 6.318, + "step": 390 + }, + { + "epoch": 0.3, + "learning_rate": 0.00022800963081861955, + "loss": 0.8193, + "step": 400 + }, + { + "epoch": 0.3, + "eval_loss": 0.7978605031967163, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.357, + "eval_steps_per_second": 6.357, + "step": 400 + }, + { + "epoch": 0.3, + "eval_loss": 0.7799659371376038, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.327, + "eval_steps_per_second": 6.327, + "step": 410 + }, + { + "epoch": 0.31, + "learning_rate": 0.00022319422150882823, + "loss": 0.8072, + "step": 420 + }, + { + "epoch": 0.31, + "eval_loss": 0.7800132036209106, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.359, + "eval_steps_per_second": 6.359, + "step": 420 + }, + { + "epoch": 0.32, + "eval_loss": 0.7845722436904907, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.329, + "eval_steps_per_second": 6.329, + "step": 430 + }, + { + "epoch": 0.33, + "learning_rate": 0.0002183788121990369, + "loss": 0.8152, + "step": 440 + }, + { + "epoch": 0.33, + "eval_loss": 0.7644073963165283, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.367, + "eval_steps_per_second": 6.367, + "step": 440 + }, + { + "epoch": 0.33, + "eval_loss": 0.7579188346862793, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.367, + "eval_steps_per_second": 6.367, + "step": 450 + }, + { + "epoch": 0.34, + "learning_rate": 0.00021356340288924557, + "loss": 0.8074, + "step": 460 + }, + { + "epoch": 0.34, + "eval_loss": 0.7676059007644653, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.341, + "eval_steps_per_second": 6.341, + "step": 460 + }, + { + "epoch": 0.35, + "eval_loss": 0.7583163976669312, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.358, + "eval_steps_per_second": 6.358, + "step": 470 + }, + { + "epoch": 0.36, + "learning_rate": 0.00020874799357945424, + "loss": 0.8118, + "step": 480 + }, + { + "epoch": 0.36, + "eval_loss": 0.7768086194992065, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.341, + "eval_steps_per_second": 6.341, + "step": 480 + }, + { + "epoch": 0.36, + "eval_loss": 0.7975252270698547, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.37, + "eval_steps_per_second": 6.37, + "step": 490 + }, + { + "epoch": 0.37, + "learning_rate": 0.0002039325842696629, + "loss": 0.8098, + "step": 500 + }, + { + "epoch": 0.37, + "eval_loss": 0.8035451173782349, + "eval_runtime": 0.1565, + "eval_samples_per_second": 6.388, + "eval_steps_per_second": 6.388, + "step": 500 + }, + { + "epoch": 0.38, + "eval_loss": 0.7718653678894043, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.347, + "eval_steps_per_second": 6.347, + "step": 510 + }, + { + "epoch": 0.39, + "learning_rate": 0.00019911717495987158, + "loss": 0.8167, + "step": 520 + }, + { + "epoch": 0.39, + "eval_loss": 0.771480917930603, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.379, + "eval_steps_per_second": 6.379, + "step": 520 + }, + { + "epoch": 0.39, + "eval_loss": 0.7757036685943604, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.364, + "eval_steps_per_second": 6.364, + "step": 530 + }, + { + "epoch": 0.4, + "learning_rate": 0.00019430176565008025, + "loss": 0.8071, + "step": 540 + }, + { + "epoch": 0.4, + "eval_loss": 0.8048112392425537, + "eval_runtime": 0.1563, + "eval_samples_per_second": 6.396, + "eval_steps_per_second": 6.396, + "step": 540 + }, + { + "epoch": 0.41, + "eval_loss": 0.7987676858901978, + "eval_runtime": 0.1562, + "eval_samples_per_second": 6.402, + "eval_steps_per_second": 6.402, + "step": 550 + }, + { + "epoch": 0.42, + "learning_rate": 0.00018948635634028892, + "loss": 0.8143, + "step": 560 + }, + { + "epoch": 0.42, + "eval_loss": 0.7960355281829834, + "eval_runtime": 0.159, + "eval_samples_per_second": 6.29, + "eval_steps_per_second": 6.29, + "step": 560 + }, + { + "epoch": 0.42, + "eval_loss": 0.7673896551132202, + "eval_runtime": 0.1569, + "eval_samples_per_second": 6.372, + "eval_steps_per_second": 6.372, + "step": 570 + }, + { + "epoch": 0.43, + "learning_rate": 0.0001846709470304976, + "loss": 0.7989, + "step": 580 + }, + { + "epoch": 0.43, + "eval_loss": 0.7655194997787476, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.336, + "eval_steps_per_second": 6.336, + "step": 580 + }, + { + "epoch": 0.44, + "eval_loss": 0.7695807218551636, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 6.365, + "step": 590 + }, + { + "epoch": 0.45, + "learning_rate": 0.00017985553772070626, + "loss": 0.8121, + "step": 600 + }, + { + "epoch": 0.45, + "eval_loss": 0.7599303722381592, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.385, + "eval_steps_per_second": 6.385, + "step": 600 + }, + { + "epoch": 0.45, + "eval_loss": 0.7409216165542603, + "eval_runtime": 0.1565, + "eval_samples_per_second": 6.389, + "eval_steps_per_second": 6.389, + "step": 610 + }, + { + "epoch": 0.46, + "learning_rate": 0.00017504012841091494, + "loss": 0.8105, + "step": 620 + }, + { + "epoch": 0.46, + "eval_loss": 0.7620519995689392, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.331, + "eval_steps_per_second": 6.331, + "step": 620 + }, + { + "epoch": 0.47, + "eval_loss": 0.7642089128494263, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.356, + "eval_steps_per_second": 6.356, + "step": 630 + }, + { + "epoch": 0.48, + "learning_rate": 0.0001702247191011236, + "loss": 0.8073, + "step": 640 + }, + { + "epoch": 0.48, + "eval_loss": 0.7464691400527954, + "eval_runtime": 0.156, + "eval_samples_per_second": 6.409, + "eval_steps_per_second": 6.409, + "step": 640 + }, + { + "epoch": 0.48, + "eval_loss": 0.7520545721054077, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.328, + "eval_steps_per_second": 6.328, + "step": 650 + }, + { + "epoch": 0.49, + "learning_rate": 0.00016540930979133222, + "loss": 0.8115, + "step": 660 + }, + { + "epoch": 0.49, + "eval_loss": 0.7851120233535767, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.358, + "eval_steps_per_second": 6.358, + "step": 660 + }, + { + "epoch": 0.5, + "eval_loss": 0.7845426797866821, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.381, + "eval_steps_per_second": 6.381, + "step": 670 + }, + { + "epoch": 0.5, + "learning_rate": 0.0001605939004815409, + "loss": 0.8213, + "step": 680 + }, + { + "epoch": 0.5, + "eval_loss": 0.7726190686225891, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.37, + "eval_steps_per_second": 6.37, + "step": 680 + }, + { + "epoch": 0.51, + "eval_loss": 0.7661857604980469, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.366, + "eval_steps_per_second": 6.366, + "step": 690 + }, + { + "epoch": 0.52, + "learning_rate": 0.00015577849117174957, + "loss": 0.8155, + "step": 700 + }, + { + "epoch": 0.52, + "eval_loss": 0.7846134305000305, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.394, + "eval_steps_per_second": 6.394, + "step": 700 + }, + { + "epoch": 0.53, + "eval_loss": 0.7888688445091248, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.324, + "eval_steps_per_second": 6.324, + "step": 710 + }, + { + "epoch": 0.53, + "learning_rate": 0.00015096308186195824, + "loss": 0.803, + "step": 720 + }, + { + "epoch": 0.53, + "eval_loss": 0.7981730103492737, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.385, + "eval_steps_per_second": 6.385, + "step": 720 + }, + { + "epoch": 0.54, + "eval_loss": 0.8129211664199829, + "eval_runtime": 0.1562, + "eval_samples_per_second": 6.401, + "eval_steps_per_second": 6.401, + "step": 730 + }, + { + "epoch": 0.55, + "learning_rate": 0.0001461476725521669, + "loss": 0.8108, + "step": 740 + }, + { + "epoch": 0.55, + "eval_loss": 0.7717241644859314, + "eval_runtime": 0.1569, + "eval_samples_per_second": 6.375, + "eval_steps_per_second": 6.375, + "step": 740 + }, + { + "epoch": 0.56, + "eval_loss": 0.7536574602127075, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.38, + "eval_steps_per_second": 6.38, + "step": 750 + }, + { + "epoch": 0.56, + "learning_rate": 0.00014133226324237558, + "loss": 0.804, + "step": 760 + }, + { + "epoch": 0.56, + "eval_loss": 0.7452749013900757, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.392, + "eval_steps_per_second": 6.392, + "step": 760 + }, + { + "epoch": 0.57, + "eval_loss": 0.7427541613578796, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.358, + "eval_steps_per_second": 6.358, + "step": 770 + }, + { + "epoch": 0.58, + "learning_rate": 0.00013651685393258425, + "loss": 0.7983, + "step": 780 + }, + { + "epoch": 0.58, + "eval_loss": 0.7469484210014343, + "eval_runtime": 0.1562, + "eval_samples_per_second": 6.404, + "eval_steps_per_second": 6.404, + "step": 780 + }, + { + "epoch": 0.59, + "eval_loss": 0.7667011022567749, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.376, + "eval_steps_per_second": 6.376, + "step": 790 + }, + { + "epoch": 0.59, + "learning_rate": 0.00013170144462279292, + "loss": 0.8054, + "step": 800 + }, + { + "epoch": 0.59, + "eval_loss": 0.7562040090560913, + "eval_runtime": 0.1651, + "eval_samples_per_second": 6.057, + "eval_steps_per_second": 6.057, + "step": 800 + }, + { + "epoch": 0.6, + "eval_loss": 0.7466799020767212, + "eval_runtime": 0.1572, + "eval_samples_per_second": 6.359, + "eval_steps_per_second": 6.359, + "step": 810 + }, + { + "epoch": 0.61, + "learning_rate": 0.0001268860353130016, + "loss": 0.8143, + "step": 820 + }, + { + "epoch": 0.61, + "eval_loss": 0.7601323127746582, + "eval_runtime": 0.1768, + "eval_samples_per_second": 5.657, + "eval_steps_per_second": 5.657, + "step": 820 + }, + { + "epoch": 0.62, + "eval_loss": 0.7430717349052429, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 830 + }, + { + "epoch": 0.62, + "learning_rate": 0.00012207062600321026, + "loss": 0.8038, + "step": 840 + }, + { + "epoch": 0.62, + "eval_loss": 0.7633467316627502, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.342, + "eval_steps_per_second": 6.342, + "step": 840 + }, + { + "epoch": 0.63, + "eval_loss": 0.7451119422912598, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 850 + }, + { + "epoch": 0.64, + "learning_rate": 0.00011725521669341892, + "loss": 0.8053, + "step": 860 + }, + { + "epoch": 0.64, + "eval_loss": 0.7628670930862427, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.338, + "eval_steps_per_second": 6.338, + "step": 860 + }, + { + "epoch": 0.65, + "eval_loss": 0.7490127086639404, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.364, + "eval_steps_per_second": 6.364, + "step": 870 + }, + { + "epoch": 0.65, + "learning_rate": 0.00011243980738362759, + "loss": 0.8007, + "step": 880 + }, + { + "epoch": 0.65, + "eval_loss": 0.767558753490448, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.392, + "eval_steps_per_second": 6.392, + "step": 880 + }, + { + "epoch": 0.66, + "eval_loss": 0.7646104097366333, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 6.365, + "step": 890 + }, + { + "epoch": 0.67, + "learning_rate": 0.00010762439807383626, + "loss": 0.7999, + "step": 900 + }, + { + "epoch": 0.67, + "eval_loss": 0.7580606937408447, + "eval_runtime": 0.1559, + "eval_samples_per_second": 6.414, + "eval_steps_per_second": 6.414, + "step": 900 + }, + { + "epoch": 0.68, + "eval_loss": 0.7789003849029541, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.377, + "eval_steps_per_second": 6.377, + "step": 910 + }, + { + "epoch": 0.68, + "learning_rate": 0.00010280898876404493, + "loss": 0.8007, + "step": 920 + }, + { + "epoch": 0.68, + "eval_loss": 0.7511205077171326, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 6.365, + "step": 920 + }, + { + "epoch": 0.69, + "eval_loss": 0.7637555599212646, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.38, + "eval_steps_per_second": 6.38, + "step": 930 + }, + { + "epoch": 0.7, + "learning_rate": 9.79935794542536e-05, + "loss": 0.8006, + "step": 940 + }, + { + "epoch": 0.7, + "eval_loss": 0.7561322450637817, + "eval_runtime": 0.1582, + "eval_samples_per_second": 6.322, + "eval_steps_per_second": 6.322, + "step": 940 + }, + { + "epoch": 0.71, + "eval_loss": 0.753495454788208, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.395, + "eval_steps_per_second": 6.395, + "step": 950 + }, + { + "epoch": 0.71, + "learning_rate": 9.317817014446228e-05, + "loss": 0.8055, + "step": 960 + }, + { + "epoch": 0.71, + "eval_loss": 0.7372075319290161, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.396, + "eval_steps_per_second": 6.396, + "step": 960 + }, + { + "epoch": 0.72, + "eval_loss": 0.761570394039154, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.329, + "eval_steps_per_second": 6.329, + "step": 970 + }, + { + "epoch": 0.73, + "learning_rate": 8.836276083467093e-05, + "loss": 0.7951, + "step": 980 + }, + { + "epoch": 0.73, + "eval_loss": 0.7430101037025452, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.386, + "eval_steps_per_second": 6.386, + "step": 980 + }, + { + "epoch": 0.74, + "eval_loss": 0.7443385124206543, + "eval_runtime": 0.1567, + "eval_samples_per_second": 6.381, + "eval_steps_per_second": 6.381, + "step": 990 + }, + { + "epoch": 0.74, + "learning_rate": 8.35473515248796e-05, + "loss": 0.8049, + "step": 1000 + }, + { + "epoch": 0.74, + "eval_loss": 0.7466243505477905, + "eval_runtime": 0.1568, + "eval_samples_per_second": 6.376, + "eval_steps_per_second": 6.376, + "step": 1000 + }, + { + "epoch": 0.75, + "eval_loss": 0.7397581934928894, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.347, + "eval_steps_per_second": 6.347, + "step": 1010 + }, + { + "epoch": 0.76, + "learning_rate": 7.873194221508827e-05, + "loss": 0.8131, + "step": 1020 + }, + { + "epoch": 0.76, + "eval_loss": 0.7271538972854614, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.35, + "eval_steps_per_second": 6.35, + "step": 1020 + }, + { + "epoch": 0.76, + "eval_loss": 0.7354034781455994, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.337, + "eval_steps_per_second": 6.337, + "step": 1030 + }, + { + "epoch": 0.77, + "learning_rate": 7.391653290529695e-05, + "loss": 0.7991, + "step": 1040 + }, + { + "epoch": 0.77, + "eval_loss": 0.7297641038894653, + "eval_runtime": 0.1764, + "eval_samples_per_second": 5.67, + "eval_steps_per_second": 5.67, + "step": 1040 + }, + { + "epoch": 0.78, + "eval_loss": 0.7396876811981201, + "eval_runtime": 0.158, + "eval_samples_per_second": 6.327, + "eval_steps_per_second": 6.327, + "step": 1050 + }, + { + "epoch": 0.79, + "learning_rate": 6.910112359550562e-05, + "loss": 0.8075, + "step": 1060 + }, + { + "epoch": 0.79, + "eval_loss": 0.7296563386917114, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.343, + "eval_steps_per_second": 6.343, + "step": 1060 + }, + { + "epoch": 0.79, + "eval_loss": 0.7264513969421387, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.366, + "eval_steps_per_second": 6.366, + "step": 1070 + }, + { + "epoch": 0.8, + "learning_rate": 6.428571428571427e-05, + "loss": 0.8075, + "step": 1080 + }, + { + "epoch": 0.8, + "eval_loss": 0.7289776802062988, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.367, + "eval_steps_per_second": 6.367, + "step": 1080 + }, + { + "epoch": 0.81, + "eval_loss": 0.7540558576583862, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 6.365, + "step": 1090 + }, + { + "epoch": 0.82, + "learning_rate": 5.9470304975922945e-05, + "loss": 0.7916, + "step": 1100 + }, + { + "epoch": 0.82, + "eval_loss": 0.7357069253921509, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.393, + "eval_steps_per_second": 6.393, + "step": 1100 + }, + { + "epoch": 0.82, + "eval_loss": 0.7496874928474426, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.351, + "eval_steps_per_second": 6.351, + "step": 1110 + }, + { + "epoch": 0.83, + "learning_rate": 5.4654895666131616e-05, + "loss": 0.7899, + "step": 1120 + }, + { + "epoch": 0.83, + "eval_loss": 0.7296478748321533, + "eval_runtime": 0.1573, + "eval_samples_per_second": 6.357, + "eval_steps_per_second": 6.357, + "step": 1120 + }, + { + "epoch": 0.84, + "eval_loss": 0.7285305261611938, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.346, + "eval_steps_per_second": 6.346, + "step": 1130 + }, + { + "epoch": 0.85, + "learning_rate": 4.983948635634029e-05, + "loss": 0.7958, + "step": 1140 + }, + { + "epoch": 0.85, + "eval_loss": 0.7548575401306152, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.368, + "eval_steps_per_second": 6.368, + "step": 1140 + }, + { + "epoch": 0.85, + "eval_loss": 0.75063556432724, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.343, + "eval_steps_per_second": 6.343, + "step": 1150 + }, + { + "epoch": 0.86, + "learning_rate": 4.502407704654895e-05, + "loss": 0.8004, + "step": 1160 + }, + { + "epoch": 0.86, + "eval_loss": 0.7478867769241333, + "eval_runtime": 0.1577, + "eval_samples_per_second": 6.341, + "eval_steps_per_second": 6.341, + "step": 1160 + }, + { + "epoch": 0.87, + "eval_loss": 0.7417320609092712, + "eval_runtime": 0.1572, + "eval_samples_per_second": 6.361, + "eval_steps_per_second": 6.361, + "step": 1170 + }, + { + "epoch": 0.88, + "learning_rate": 4.020866773675762e-05, + "loss": 0.7927, + "step": 1180 + }, + { + "epoch": 0.88, + "eval_loss": 0.7448097467422485, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.337, + "eval_steps_per_second": 6.337, + "step": 1180 + }, + { + "epoch": 0.88, + "eval_loss": 0.7570788860321045, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.346, + "eval_steps_per_second": 6.346, + "step": 1190 + }, + { + "epoch": 0.89, + "learning_rate": 3.539325842696629e-05, + "loss": 0.8013, + "step": 1200 + }, + { + "epoch": 0.89, + "eval_loss": 0.7594838738441467, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.324, + "eval_steps_per_second": 6.324, + "step": 1200 + }, + { + "epoch": 0.9, + "eval_loss": 0.7382398843765259, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.369, + "eval_steps_per_second": 6.369, + "step": 1210 + }, + { + "epoch": 0.91, + "learning_rate": 3.057784911717496e-05, + "loss": 0.8002, + "step": 1220 + }, + { + "epoch": 0.91, + "eval_loss": 0.7314504384994507, + "eval_runtime": 0.1571, + "eval_samples_per_second": 6.366, + "eval_steps_per_second": 6.366, + "step": 1220 + }, + { + "epoch": 0.91, + "eval_loss": 0.780935525894165, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.348, + "eval_steps_per_second": 6.348, + "step": 1230 + }, + { + "epoch": 0.92, + "learning_rate": 2.5762439807383626e-05, + "loss": 0.7914, + "step": 1240 + }, + { + "epoch": 0.92, + "eval_loss": 0.7425116300582886, + "eval_runtime": 0.1574, + "eval_samples_per_second": 6.352, + "eval_steps_per_second": 6.352, + "step": 1240 + }, + { + "epoch": 0.93, + "eval_loss": 0.7363260984420776, + "eval_runtime": 0.1564, + "eval_samples_per_second": 6.394, + "eval_steps_per_second": 6.394, + "step": 1250 + }, + { + "epoch": 0.94, + "learning_rate": 2.0947030497592293e-05, + "loss": 0.7911, + "step": 1260 + }, + { + "epoch": 0.94, + "eval_loss": 0.7363616228103638, + "eval_runtime": 0.1566, + "eval_samples_per_second": 6.385, + "eval_steps_per_second": 6.385, + "step": 1260 + }, + { + "epoch": 0.94, + "eval_loss": 0.7710204124450684, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 1270 + }, + { + "epoch": 0.95, + "learning_rate": 1.613162118780096e-05, + "loss": 0.7973, + "step": 1280 + }, + { + "epoch": 0.95, + "eval_loss": 0.7362488508224487, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.349, + "eval_steps_per_second": 6.349, + "step": 1280 + }, + { + "epoch": 0.96, + "eval_loss": 0.7484357357025146, + "eval_runtime": 0.1581, + "eval_samples_per_second": 6.326, + "eval_steps_per_second": 6.326, + "step": 1290 + }, + { + "epoch": 0.97, + "learning_rate": 1.1316211878009629e-05, + "loss": 0.8008, + "step": 1300 + }, + { + "epoch": 0.97, + "eval_loss": 0.7393674254417419, + "eval_runtime": 0.1578, + "eval_samples_per_second": 6.338, + "eval_steps_per_second": 6.338, + "step": 1300 + }, + { + "epoch": 0.97, + "eval_loss": 0.738249659538269, + "eval_runtime": 0.1576, + "eval_samples_per_second": 6.344, + "eval_steps_per_second": 6.344, + "step": 1310 + }, + { + "epoch": 0.98, + "learning_rate": 6.500802568218298e-06, + "loss": 0.798, + "step": 1320 + }, + { + "epoch": 0.98, + "eval_loss": 0.7329107522964478, + "eval_runtime": 0.1575, + "eval_samples_per_second": 6.351, + "eval_steps_per_second": 6.351, + "step": 1320 + }, + { + "epoch": 0.99, + "eval_loss": 0.7391780614852905, + "eval_runtime": 0.1587, + "eval_samples_per_second": 6.302, + "eval_steps_per_second": 6.302, + "step": 1330 + }, + { + "epoch": 0.99, + "learning_rate": 1.6853932584269661e-06, + "loss": 0.7902, + "step": 1340 + }, + { + "epoch": 0.99, + "eval_loss": 0.7421530485153198, + "eval_runtime": 0.157, + "eval_samples_per_second": 6.37, + "eval_steps_per_second": 6.37, + "step": 1340 + } + ], + "max_steps": 1346, + "num_train_epochs": 1, + "total_flos": 1.7418408714829824e+18, + "trial_name": null, + "trial_params": null +} diff --git a/llama/lora-alpaca-q90-use-seq/checkpoint-1340/training_args.bin b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..2ca60f7ff946be4d5dd08bedeca15733e2f014af --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/checkpoint-1340/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b92cf17cb7884e6a8a0c6e72287125f12b4939d9d31cda3054f19a39cab8a0f +size 3579 diff --git a/llama/lora-alpaca-q90-use-seq/runs/Apr05_22-06-44_walnut/1680732406.673894/events.out.tfevents.1680732406.walnut.45965.1 b/llama/lora-alpaca-q90-use-seq/runs/Apr05_22-06-44_walnut/1680732406.673894/events.out.tfevents.1680732406.walnut.45965.1 new file mode 100644 index 0000000000000000000000000000000000000000..859dc5c8c7bd6e68707108e5232cd04c67fb0113 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/runs/Apr05_22-06-44_walnut/1680732406.673894/events.out.tfevents.1680732406.walnut.45965.1 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:974df580cdc67859d445c2fbbcbe523526565737c3555b999708052c6006dfaf +size 5830 diff --git a/llama/lora-alpaca-q90-use-seq/runs/Apr05_22-06-44_walnut/events.out.tfevents.1680732406.walnut.45965.0 b/llama/lora-alpaca-q90-use-seq/runs/Apr05_22-06-44_walnut/events.out.tfevents.1680732406.walnut.45965.0 new file mode 100644 index 0000000000000000000000000000000000000000..abfa61e41a9b7641c819401bcbea518bef0da8e6 --- /dev/null +++ b/llama/lora-alpaca-q90-use-seq/runs/Apr05_22-06-44_walnut/events.out.tfevents.1680732406.walnut.45965.0 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0ca15878fa655be5d52f1f173a40984848e0b7c3224c01d74806c2b7a629cf13 +size 51022