Instructions to use craa/exceptions with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use craa/exceptions with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="craa/exceptions")# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("craa/exceptions", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use craa/exceptions with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "craa/exceptions" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "craa/exceptions", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/craa/exceptions
- SGLang
How to use craa/exceptions with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "craa/exceptions" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "craa/exceptions", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "craa/exceptions" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "craa/exceptions", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use craa/exceptions with Docker Model Runner:
docker model run hf.co/craa/exceptions
Download exp3/high_2000_6910/trainer_state.json from craa/exceptions: direct link, hf CLI and curl.
- Browser
- Download file 411 kB
-
https://huggingface.co/craa/exceptions/resolve/main/exp3/high_2000_6910/trainer_state.json
- Command line
-
hf download hf://craa/exceptions/exp3/high_2000_6910/trainer_state.json
-
curl -L -o trainer_state.json https://huggingface.co/craa/exceptions/resolve/main/exp3/high_2000_6910/trainer_state.json
411 kB
{
"best_global_step": 89000,
"best_metric": 3.275252103805542,
"best_model_checkpoint": "/scratch/cl5625/exceptions/models/high_2000_6910/checkpoint-70000",
"epoch": 29.33268733850129,
"eval_steps": 1000,
"global_step": 109000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.013458225667527994,
"grad_norm": 3.308773994445801,
"learning_rate": 0.000294,
"loss": 8.4679,
"step": 50
},
{
"epoch": 0.026916451335055987,
"grad_norm": 1.6042652130126953,
"learning_rate": 0.0005939999999999999,
"loss": 6.7351,
"step": 100
},
{
"epoch": 0.04037467700258398,
"grad_norm": 1.6990554332733154,
"learning_rate": 0.0005998020735155513,
"loss": 6.3185,
"step": 150
},
{
"epoch": 0.053832902670111975,
"grad_norm": 0.9088912010192871,
"learning_rate": 0.0005996001077150935,
"loss": 6.0464,
"step": 200
},
{
"epoch": 0.06729112833763996,
"grad_norm": 0.6608636379241943,
"learning_rate": 0.0005993981419146358,
"loss": 5.8696,
"step": 250
},
{
"epoch": 0.08074935400516796,
"grad_norm": 1.4318311214447021,
"learning_rate": 0.0005991961761141779,
"loss": 5.7427,
"step": 300
},
{
"epoch": 0.09420757967269595,
"grad_norm": 2.375964641571045,
"learning_rate": 0.0005989942103137202,
"loss": 5.6263,
"step": 350
},
{
"epoch": 0.10766580534022395,
"grad_norm": 1.0631674528121948,
"learning_rate": 0.0005987922445132624,
"loss": 5.4973,
"step": 400
},
{
"epoch": 0.12112403100775193,
"grad_norm": 1.3984321355819702,
"learning_rate": 0.0005985902787128047,
"loss": 5.3815,
"step": 450
},
{
"epoch": 0.13458225667527993,
"grad_norm": 1.0844402313232422,
"learning_rate": 0.0005983883129123468,
"loss": 5.2436,
"step": 500
},
{
"epoch": 0.14804048234280792,
"grad_norm": 1.4188182353973389,
"learning_rate": 0.000598186347111889,
"loss": 5.1461,
"step": 550
},
{
"epoch": 0.16149870801033592,
"grad_norm": 1.540555715560913,
"learning_rate": 0.0005979843813114312,
"loss": 5.0629,
"step": 600
},
{
"epoch": 0.1749569336778639,
"grad_norm": 1.31498122215271,
"learning_rate": 0.0005977824155109734,
"loss": 4.9972,
"step": 650
},
{
"epoch": 0.1884151593453919,
"grad_norm": 1.060059666633606,
"learning_rate": 0.0005975804497105157,
"loss": 4.9156,
"step": 700
},
{
"epoch": 0.2018733850129199,
"grad_norm": 0.7784032821655273,
"learning_rate": 0.0005973784839100578,
"loss": 4.8646,
"step": 750
},
{
"epoch": 0.2153316106804479,
"grad_norm": 1.025396466255188,
"learning_rate": 0.0005971765181096001,
"loss": 4.8039,
"step": 800
},
{
"epoch": 0.2287898363479759,
"grad_norm": 0.7960473895072937,
"learning_rate": 0.0005969745523091422,
"loss": 4.7349,
"step": 850
},
{
"epoch": 0.24224806201550386,
"grad_norm": 0.9397687911987305,
"learning_rate": 0.0005967725865086845,
"loss": 4.6972,
"step": 900
},
{
"epoch": 0.2557062876830319,
"grad_norm": 0.9388722777366638,
"learning_rate": 0.0005965706207082267,
"loss": 4.6541,
"step": 950
},
{
"epoch": 0.26916451335055985,
"grad_norm": 0.8853117227554321,
"learning_rate": 0.0005963686549077689,
"loss": 4.6012,
"step": 1000
},
{
"epoch": 0.26916451335055985,
"eval_accuracy": 0.2658181129953373,
"eval_loss": 4.529858589172363,
"eval_runtime": 52.8669,
"eval_samples_per_second": 340.705,
"eval_steps_per_second": 21.299,
"step": 1000
},
{
"epoch": 0.2826227390180879,
"grad_norm": 1.1518616676330566,
"learning_rate": 0.0005961666891073111,
"loss": 4.5541,
"step": 1050
},
{
"epoch": 0.29608096468561584,
"grad_norm": 0.9574114680290222,
"learning_rate": 0.0005959647233068533,
"loss": 4.4977,
"step": 1100
},
{
"epoch": 0.30953919035314387,
"grad_norm": 1.084009051322937,
"learning_rate": 0.0005957627575063955,
"loss": 4.4619,
"step": 1150
},
{
"epoch": 0.32299741602067183,
"grad_norm": 0.7612378001213074,
"learning_rate": 0.0005955607917059378,
"loss": 4.4404,
"step": 1200
},
{
"epoch": 0.3364556416881998,
"grad_norm": 0.8939399719238281,
"learning_rate": 0.00059535882590548,
"loss": 4.3977,
"step": 1250
},
{
"epoch": 0.3499138673557278,
"grad_norm": 0.7879042625427246,
"learning_rate": 0.0005951568601050221,
"loss": 4.3644,
"step": 1300
},
{
"epoch": 0.3633720930232558,
"grad_norm": 0.9349780082702637,
"learning_rate": 0.0005949548943045644,
"loss": 4.3304,
"step": 1350
},
{
"epoch": 0.3768303186907838,
"grad_norm": 0.9910395741462708,
"learning_rate": 0.0005947529285041066,
"loss": 4.3176,
"step": 1400
},
{
"epoch": 0.3902885443583118,
"grad_norm": 0.915924608707428,
"learning_rate": 0.0005945509627036488,
"loss": 4.279,
"step": 1450
},
{
"epoch": 0.4037467700258398,
"grad_norm": 0.8442095518112183,
"learning_rate": 0.000594348996903191,
"loss": 4.2577,
"step": 1500
},
{
"epoch": 0.4172049956933678,
"grad_norm": 0.7409194707870483,
"learning_rate": 0.0005941470311027332,
"loss": 4.233,
"step": 1550
},
{
"epoch": 0.4306632213608958,
"grad_norm": 0.6870989799499512,
"learning_rate": 0.0005939450653022754,
"loss": 4.2171,
"step": 1600
},
{
"epoch": 0.44412144702842377,
"grad_norm": 1.060454249382019,
"learning_rate": 0.0005937430995018177,
"loss": 4.1993,
"step": 1650
},
{
"epoch": 0.4575796726959518,
"grad_norm": 1.0149903297424316,
"learning_rate": 0.0005935411337013598,
"loss": 4.1869,
"step": 1700
},
{
"epoch": 0.47103789836347976,
"grad_norm": 0.7430284023284912,
"learning_rate": 0.0005933391679009021,
"loss": 4.163,
"step": 1750
},
{
"epoch": 0.4844961240310077,
"grad_norm": 0.5982431769371033,
"learning_rate": 0.0005931372021004442,
"loss": 4.1555,
"step": 1800
},
{
"epoch": 0.49795434969853575,
"grad_norm": 0.708482563495636,
"learning_rate": 0.0005929352362999865,
"loss": 4.1336,
"step": 1850
},
{
"epoch": 0.5114125753660638,
"grad_norm": 0.8102256655693054,
"learning_rate": 0.0005927332704995287,
"loss": 4.1256,
"step": 1900
},
{
"epoch": 0.5248708010335917,
"grad_norm": 0.6787132024765015,
"learning_rate": 0.0005925313046990709,
"loss": 4.1054,
"step": 1950
},
{
"epoch": 0.5383290267011197,
"grad_norm": 0.6520165801048279,
"learning_rate": 0.0005923293388986131,
"loss": 4.085,
"step": 2000
},
{
"epoch": 0.5383290267011197,
"eval_accuracy": 0.31672122911753114,
"eval_loss": 4.02903413772583,
"eval_runtime": 53.0678,
"eval_samples_per_second": 339.415,
"eval_steps_per_second": 21.218,
"step": 2000
},
{
"epoch": 0.5517872523686477,
"grad_norm": 0.6289111971855164,
"learning_rate": 0.0005921273730981553,
"loss": 4.0798,
"step": 2050
},
{
"epoch": 0.5652454780361758,
"grad_norm": 0.6548891663551331,
"learning_rate": 0.0005919254072976975,
"loss": 4.056,
"step": 2100
},
{
"epoch": 0.5787037037037037,
"grad_norm": 0.6513845920562744,
"learning_rate": 0.0005917234414972398,
"loss": 4.0544,
"step": 2150
},
{
"epoch": 0.5921619293712317,
"grad_norm": 0.5987239480018616,
"learning_rate": 0.000591521475696782,
"loss": 4.0385,
"step": 2200
},
{
"epoch": 0.6056201550387597,
"grad_norm": 0.5677440762519836,
"learning_rate": 0.0005913195098963241,
"loss": 4.0397,
"step": 2250
},
{
"epoch": 0.6190783807062877,
"grad_norm": 0.5570490956306458,
"learning_rate": 0.0005911175440958664,
"loss": 4.0091,
"step": 2300
},
{
"epoch": 0.6325366063738157,
"grad_norm": 0.5314527750015259,
"learning_rate": 0.0005909155782954086,
"loss": 4.0057,
"step": 2350
},
{
"epoch": 0.6459948320413437,
"grad_norm": 0.7578822374343872,
"learning_rate": 0.0005907136124949508,
"loss": 4.0038,
"step": 2400
},
{
"epoch": 0.6594530577088716,
"grad_norm": 0.8584759831428528,
"learning_rate": 0.000590511646694493,
"loss": 3.994,
"step": 2450
},
{
"epoch": 0.6729112833763996,
"grad_norm": 0.5055056214332581,
"learning_rate": 0.0005903096808940352,
"loss": 3.9828,
"step": 2500
},
{
"epoch": 0.6863695090439277,
"grad_norm": 0.7988194823265076,
"learning_rate": 0.0005901077150935774,
"loss": 3.9744,
"step": 2550
},
{
"epoch": 0.6998277347114557,
"grad_norm": 0.5637562870979309,
"learning_rate": 0.0005899057492931197,
"loss": 3.9678,
"step": 2600
},
{
"epoch": 0.7132859603789836,
"grad_norm": 0.5229179859161377,
"learning_rate": 0.0005897037834926618,
"loss": 3.9523,
"step": 2650
},
{
"epoch": 0.7267441860465116,
"grad_norm": 0.588083803653717,
"learning_rate": 0.0005895018176922041,
"loss": 3.9488,
"step": 2700
},
{
"epoch": 0.7402024117140397,
"grad_norm": 0.6086938977241516,
"learning_rate": 0.0005892998518917462,
"loss": 3.9542,
"step": 2750
},
{
"epoch": 0.7536606373815676,
"grad_norm": 0.5917510986328125,
"learning_rate": 0.0005890978860912885,
"loss": 3.9318,
"step": 2800
},
{
"epoch": 0.7671188630490956,
"grad_norm": 0.5497215390205383,
"learning_rate": 0.0005888959202908307,
"loss": 3.9295,
"step": 2850
},
{
"epoch": 0.7805770887166236,
"grad_norm": 0.4543793499469757,
"learning_rate": 0.0005886939544903729,
"loss": 3.9134,
"step": 2900
},
{
"epoch": 0.7940353143841516,
"grad_norm": 0.6304684281349182,
"learning_rate": 0.0005884919886899151,
"loss": 3.9112,
"step": 2950
},
{
"epoch": 0.8074935400516796,
"grad_norm": 0.6081494688987732,
"learning_rate": 0.0005882900228894573,
"loss": 3.9068,
"step": 3000
},
{
"epoch": 0.8074935400516796,
"eval_accuracy": 0.3333811379497817,
"eval_loss": 3.844052791595459,
"eval_runtime": 53.1523,
"eval_samples_per_second": 338.875,
"eval_steps_per_second": 21.184,
"step": 3000
},
{
"epoch": 0.8209517657192076,
"grad_norm": 0.5118460655212402,
"learning_rate": 0.0005880880570889996,
"loss": 3.8916,
"step": 3050
},
{
"epoch": 0.8344099913867356,
"grad_norm": 0.5820624232292175,
"learning_rate": 0.0005878860912885418,
"loss": 3.881,
"step": 3100
},
{
"epoch": 0.8478682170542635,
"grad_norm": 0.5204915404319763,
"learning_rate": 0.000587684125488084,
"loss": 3.8647,
"step": 3150
},
{
"epoch": 0.8613264427217916,
"grad_norm": 0.4955189824104309,
"learning_rate": 0.0005874821596876261,
"loss": 3.8708,
"step": 3200
},
{
"epoch": 0.8747846683893196,
"grad_norm": 0.5090410709381104,
"learning_rate": 0.0005872801938871684,
"loss": 3.863,
"step": 3250
},
{
"epoch": 0.8882428940568475,
"grad_norm": 0.48452430963516235,
"learning_rate": 0.0005870782280867106,
"loss": 3.8626,
"step": 3300
},
{
"epoch": 0.9017011197243755,
"grad_norm": 0.49206408858299255,
"learning_rate": 0.0005868762622862528,
"loss": 3.8704,
"step": 3350
},
{
"epoch": 0.9151593453919036,
"grad_norm": 0.5284396409988403,
"learning_rate": 0.000586674296485795,
"loss": 3.8523,
"step": 3400
},
{
"epoch": 0.9286175710594315,
"grad_norm": 0.44494983553886414,
"learning_rate": 0.0005864723306853373,
"loss": 3.8304,
"step": 3450
},
{
"epoch": 0.9420757967269595,
"grad_norm": 0.4655598998069763,
"learning_rate": 0.0005862703648848794,
"loss": 3.837,
"step": 3500
},
{
"epoch": 0.9555340223944875,
"grad_norm": 0.4665937125682831,
"learning_rate": 0.0005860683990844217,
"loss": 3.8193,
"step": 3550
},
{
"epoch": 0.9689922480620154,
"grad_norm": 0.4261581599712372,
"learning_rate": 0.0005858664332839638,
"loss": 3.8203,
"step": 3600
},
{
"epoch": 0.9824504737295435,
"grad_norm": 0.5406307578086853,
"learning_rate": 0.0005856644674835061,
"loss": 3.8193,
"step": 3650
},
{
"epoch": 0.9959086993970715,
"grad_norm": 0.47078415751457214,
"learning_rate": 0.0005854625016830483,
"loss": 3.8235,
"step": 3700
},
{
"epoch": 1.009151593453919,
"grad_norm": 0.44889238476753235,
"learning_rate": 0.0005852605358825905,
"loss": 3.7595,
"step": 3750
},
{
"epoch": 1.022609819121447,
"grad_norm": 0.4847916066646576,
"learning_rate": 0.0005850585700821327,
"loss": 3.7577,
"step": 3800
},
{
"epoch": 1.036068044788975,
"grad_norm": 0.4172884523868561,
"learning_rate": 0.0005848566042816749,
"loss": 3.7459,
"step": 3850
},
{
"epoch": 1.049526270456503,
"grad_norm": 0.42249923944473267,
"learning_rate": 0.0005846546384812171,
"loss": 3.7445,
"step": 3900
},
{
"epoch": 1.062984496124031,
"grad_norm": 0.4443018436431885,
"learning_rate": 0.0005844526726807594,
"loss": 3.7393,
"step": 3950
},
{
"epoch": 1.076442721791559,
"grad_norm": 0.49296411871910095,
"learning_rate": 0.0005842507068803016,
"loss": 3.7396,
"step": 4000
},
{
"epoch": 1.076442721791559,
"eval_accuracy": 0.3439660578531468,
"eval_loss": 3.736942768096924,
"eval_runtime": 53.1277,
"eval_samples_per_second": 339.032,
"eval_steps_per_second": 21.194,
"step": 4000
},
{
"epoch": 1.089900947459087,
"grad_norm": 0.4840051531791687,
"learning_rate": 0.0005840487410798437,
"loss": 3.7276,
"step": 4050
},
{
"epoch": 1.103359173126615,
"grad_norm": 0.44972163438796997,
"learning_rate": 0.000583846775279386,
"loss": 3.7509,
"step": 4100
},
{
"epoch": 1.116817398794143,
"grad_norm": 0.5025952458381653,
"learning_rate": 0.0005836448094789282,
"loss": 3.7371,
"step": 4150
},
{
"epoch": 1.130275624461671,
"grad_norm": 0.49098411202430725,
"learning_rate": 0.0005834428436784704,
"loss": 3.7385,
"step": 4200
},
{
"epoch": 1.143733850129199,
"grad_norm": 0.4680483639240265,
"learning_rate": 0.0005832408778780126,
"loss": 3.7186,
"step": 4250
},
{
"epoch": 1.157192075796727,
"grad_norm": 0.43116289377212524,
"learning_rate": 0.0005830389120775548,
"loss": 3.7223,
"step": 4300
},
{
"epoch": 1.1706503014642549,
"grad_norm": 0.46523234248161316,
"learning_rate": 0.000582836946277097,
"loss": 3.7236,
"step": 4350
},
{
"epoch": 1.1841085271317828,
"grad_norm": 0.41721317172050476,
"learning_rate": 0.0005826349804766393,
"loss": 3.723,
"step": 4400
},
{
"epoch": 1.197566752799311,
"grad_norm": 0.4401717782020569,
"learning_rate": 0.0005824330146761814,
"loss": 3.7283,
"step": 4450
},
{
"epoch": 1.211024978466839,
"grad_norm": 0.43926772475242615,
"learning_rate": 0.0005822310488757237,
"loss": 3.7224,
"step": 4500
},
{
"epoch": 1.224483204134367,
"grad_norm": 0.43538621068000793,
"learning_rate": 0.0005820290830752658,
"loss": 3.7151,
"step": 4550
},
{
"epoch": 1.237941429801895,
"grad_norm": 0.3874562978744507,
"learning_rate": 0.0005818271172748081,
"loss": 3.7128,
"step": 4600
},
{
"epoch": 1.251399655469423,
"grad_norm": 0.4349665343761444,
"learning_rate": 0.0005816251514743503,
"loss": 3.7068,
"step": 4650
},
{
"epoch": 1.2648578811369509,
"grad_norm": 0.4198451638221741,
"learning_rate": 0.0005814231856738925,
"loss": 3.7007,
"step": 4700
},
{
"epoch": 1.2783161068044788,
"grad_norm": 0.45340976119041443,
"learning_rate": 0.0005812212198734347,
"loss": 3.7071,
"step": 4750
},
{
"epoch": 1.291774332472007,
"grad_norm": 0.41575804352760315,
"learning_rate": 0.0005810192540729769,
"loss": 3.7074,
"step": 4800
},
{
"epoch": 1.3052325581395348,
"grad_norm": 0.5499830842018127,
"learning_rate": 0.0005808172882725192,
"loss": 3.7059,
"step": 4850
},
{
"epoch": 1.318690783807063,
"grad_norm": 0.42270398139953613,
"learning_rate": 0.0005806153224720614,
"loss": 3.7028,
"step": 4900
},
{
"epoch": 1.332149009474591,
"grad_norm": 0.41370531916618347,
"learning_rate": 0.0005804133566716036,
"loss": 3.6946,
"step": 4950
},
{
"epoch": 1.3456072351421189,
"grad_norm": 0.4177495539188385,
"learning_rate": 0.0005802113908711457,
"loss": 3.6861,
"step": 5000
},
{
"epoch": 1.3456072351421189,
"eval_accuracy": 0.35080939734458394,
"eval_loss": 3.6666910648345947,
"eval_runtime": 53.2598,
"eval_samples_per_second": 338.191,
"eval_steps_per_second": 21.142,
"step": 5000
},
{
"epoch": 1.3590654608096469,
"grad_norm": 0.3675026595592499,
"learning_rate": 0.000580009425070688,
"loss": 3.6914,
"step": 5050
},
{
"epoch": 1.3725236864771748,
"grad_norm": 0.41109681129455566,
"learning_rate": 0.0005798074592702302,
"loss": 3.6908,
"step": 5100
},
{
"epoch": 1.3859819121447028,
"grad_norm": 0.4249870181083679,
"learning_rate": 0.0005796054934697724,
"loss": 3.6828,
"step": 5150
},
{
"epoch": 1.3994401378122308,
"grad_norm": 0.4249998927116394,
"learning_rate": 0.0005794035276693146,
"loss": 3.6723,
"step": 5200
},
{
"epoch": 1.412898363479759,
"grad_norm": 0.4438369572162628,
"learning_rate": 0.0005792015618688568,
"loss": 3.6659,
"step": 5250
},
{
"epoch": 1.4263565891472867,
"grad_norm": 0.48220908641815186,
"learning_rate": 0.000578999596068399,
"loss": 3.6638,
"step": 5300
},
{
"epoch": 1.4398148148148149,
"grad_norm": 0.4296906888484955,
"learning_rate": 0.0005787976302679413,
"loss": 3.6744,
"step": 5350
},
{
"epoch": 1.4532730404823428,
"grad_norm": 0.4431769847869873,
"learning_rate": 0.0005785956644674834,
"loss": 3.6551,
"step": 5400
},
{
"epoch": 1.4667312661498708,
"grad_norm": 0.41457149386405945,
"learning_rate": 0.0005783936986670257,
"loss": 3.6671,
"step": 5450
},
{
"epoch": 1.4801894918173988,
"grad_norm": 0.39893487095832825,
"learning_rate": 0.0005781917328665678,
"loss": 3.6632,
"step": 5500
},
{
"epoch": 1.4936477174849268,
"grad_norm": 0.3814569413661957,
"learning_rate": 0.0005779897670661101,
"loss": 3.6514,
"step": 5550
},
{
"epoch": 1.507105943152455,
"grad_norm": 0.39305347204208374,
"learning_rate": 0.0005777878012656523,
"loss": 3.652,
"step": 5600
},
{
"epoch": 1.5205641688199827,
"grad_norm": 0.4522084891796112,
"learning_rate": 0.0005775858354651945,
"loss": 3.6533,
"step": 5650
},
{
"epoch": 1.5340223944875109,
"grad_norm": 0.3905489444732666,
"learning_rate": 0.0005773838696647367,
"loss": 3.6526,
"step": 5700
},
{
"epoch": 1.5474806201550386,
"grad_norm": 0.3704451620578766,
"learning_rate": 0.000577181903864279,
"loss": 3.647,
"step": 5750
},
{
"epoch": 1.5609388458225668,
"grad_norm": 0.38568341732025146,
"learning_rate": 0.0005769799380638212,
"loss": 3.6492,
"step": 5800
},
{
"epoch": 1.5743970714900948,
"grad_norm": 0.45711877942085266,
"learning_rate": 0.0005767779722633633,
"loss": 3.6387,
"step": 5850
},
{
"epoch": 1.5878552971576227,
"grad_norm": 0.41149041056632996,
"learning_rate": 0.0005765760064629056,
"loss": 3.6423,
"step": 5900
},
{
"epoch": 1.6013135228251507,
"grad_norm": 0.39101529121398926,
"learning_rate": 0.0005763740406624477,
"loss": 3.6335,
"step": 5950
},
{
"epoch": 1.6147717484926787,
"grad_norm": 0.4096600413322449,
"learning_rate": 0.00057617207486199,
"loss": 3.634,
"step": 6000
},
{
"epoch": 1.6147717484926787,
"eval_accuracy": 0.35582301514146036,
"eval_loss": 3.612297534942627,
"eval_runtime": 53.361,
"eval_samples_per_second": 337.55,
"eval_steps_per_second": 21.102,
"step": 6000
},
{
"epoch": 1.6282299741602069,
"grad_norm": 0.3830372989177704,
"learning_rate": 0.0005759701090615322,
"loss": 3.6409,
"step": 6050
},
{
"epoch": 1.6416881998277346,
"grad_norm": 0.3905888497829437,
"learning_rate": 0.0005757681432610744,
"loss": 3.6282,
"step": 6100
},
{
"epoch": 1.6551464254952628,
"grad_norm": 0.3588292598724365,
"learning_rate": 0.0005755661774606166,
"loss": 3.6315,
"step": 6150
},
{
"epoch": 1.6686046511627906,
"grad_norm": 0.39833560585975647,
"learning_rate": 0.0005753642116601588,
"loss": 3.6292,
"step": 6200
},
{
"epoch": 1.6820628768303187,
"grad_norm": 0.38218745589256287,
"learning_rate": 0.000575162245859701,
"loss": 3.6251,
"step": 6250
},
{
"epoch": 1.6955211024978467,
"grad_norm": 0.35812291502952576,
"learning_rate": 0.0005749602800592433,
"loss": 3.635,
"step": 6300
},
{
"epoch": 1.7089793281653747,
"grad_norm": 0.4691900312900543,
"learning_rate": 0.0005747583142587854,
"loss": 3.6139,
"step": 6350
},
{
"epoch": 1.7224375538329026,
"grad_norm": 0.37986138463020325,
"learning_rate": 0.0005745563484583277,
"loss": 3.6301,
"step": 6400
},
{
"epoch": 1.7358957795004306,
"grad_norm": 0.38206204771995544,
"learning_rate": 0.0005743543826578698,
"loss": 3.6196,
"step": 6450
},
{
"epoch": 1.7493540051679588,
"grad_norm": 0.37841781973838806,
"learning_rate": 0.0005741524168574121,
"loss": 3.6116,
"step": 6500
},
{
"epoch": 1.7628122308354865,
"grad_norm": 0.381045401096344,
"learning_rate": 0.0005739504510569543,
"loss": 3.611,
"step": 6550
},
{
"epoch": 1.7762704565030147,
"grad_norm": 0.4393485188484192,
"learning_rate": 0.0005737484852564966,
"loss": 3.6084,
"step": 6600
},
{
"epoch": 1.7897286821705425,
"grad_norm": 0.36671558022499084,
"learning_rate": 0.0005735465194560387,
"loss": 3.616,
"step": 6650
},
{
"epoch": 1.8031869078380707,
"grad_norm": 0.3708931505680084,
"learning_rate": 0.000573344553655581,
"loss": 3.5929,
"step": 6700
},
{
"epoch": 1.8166451335055986,
"grad_norm": 0.3850688338279724,
"learning_rate": 0.0005731425878551232,
"loss": 3.6034,
"step": 6750
},
{
"epoch": 1.8301033591731266,
"grad_norm": 0.39316853880882263,
"learning_rate": 0.0005729406220546653,
"loss": 3.6017,
"step": 6800
},
{
"epoch": 1.8435615848406546,
"grad_norm": 0.4011707007884979,
"learning_rate": 0.0005727386562542076,
"loss": 3.6086,
"step": 6850
},
{
"epoch": 1.8570198105081825,
"grad_norm": 0.4198000133037567,
"learning_rate": 0.0005725366904537497,
"loss": 3.6026,
"step": 6900
},
{
"epoch": 1.8704780361757107,
"grad_norm": 0.385288804769516,
"learning_rate": 0.000572334724653292,
"loss": 3.6097,
"step": 6950
},
{
"epoch": 1.8839362618432385,
"grad_norm": 0.4373071789741516,
"learning_rate": 0.0005721327588528342,
"loss": 3.602,
"step": 7000
},
{
"epoch": 1.8839362618432385,
"eval_accuracy": 0.36056349474344784,
"eval_loss": 3.567089319229126,
"eval_runtime": 53.2558,
"eval_samples_per_second": 338.217,
"eval_steps_per_second": 21.143,
"step": 7000
},
{
"epoch": 1.8973944875107667,
"grad_norm": 0.4110179841518402,
"learning_rate": 0.0005719307930523764,
"loss": 3.5933,
"step": 7050
},
{
"epoch": 1.9108527131782944,
"grad_norm": 0.3766134977340698,
"learning_rate": 0.0005717288272519186,
"loss": 3.6057,
"step": 7100
},
{
"epoch": 1.9243109388458226,
"grad_norm": 0.4383994936943054,
"learning_rate": 0.0005715268614514608,
"loss": 3.6037,
"step": 7150
},
{
"epoch": 1.9377691645133506,
"grad_norm": 0.38433584570884705,
"learning_rate": 0.000571324895651003,
"loss": 3.5909,
"step": 7200
},
{
"epoch": 1.9512273901808785,
"grad_norm": 0.3365281820297241,
"learning_rate": 0.0005711229298505453,
"loss": 3.5885,
"step": 7250
},
{
"epoch": 1.9646856158484065,
"grad_norm": 0.39374086260795593,
"learning_rate": 0.0005709209640500874,
"loss": 3.5944,
"step": 7300
},
{
"epoch": 1.9781438415159345,
"grad_norm": 0.37078049778938293,
"learning_rate": 0.0005707189982496297,
"loss": 3.5874,
"step": 7350
},
{
"epoch": 1.9916020671834627,
"grad_norm": 0.42410802841186523,
"learning_rate": 0.0005705170324491719,
"loss": 3.5836,
"step": 7400
},
{
"epoch": 2.00484496124031,
"grad_norm": 0.3713061213493347,
"learning_rate": 0.0005703150666487141,
"loss": 3.5666,
"step": 7450
},
{
"epoch": 2.018303186907838,
"grad_norm": 0.37857165932655334,
"learning_rate": 0.0005701131008482563,
"loss": 3.5089,
"step": 7500
},
{
"epoch": 2.0317614125753662,
"grad_norm": 0.3648632764816284,
"learning_rate": 0.0005699111350477986,
"loss": 3.4958,
"step": 7550
},
{
"epoch": 2.045219638242894,
"grad_norm": 0.35465335845947266,
"learning_rate": 0.0005697091692473408,
"loss": 3.5056,
"step": 7600
},
{
"epoch": 2.058677863910422,
"grad_norm": 0.38645628094673157,
"learning_rate": 0.000569507203446883,
"loss": 3.5053,
"step": 7650
},
{
"epoch": 2.07213608957795,
"grad_norm": 0.3809046149253845,
"learning_rate": 0.0005693052376464252,
"loss": 3.5084,
"step": 7700
},
{
"epoch": 2.085594315245478,
"grad_norm": 0.41608577966690063,
"learning_rate": 0.0005691032718459673,
"loss": 3.5104,
"step": 7750
},
{
"epoch": 2.099052540913006,
"grad_norm": 0.3553376793861389,
"learning_rate": 0.0005689013060455096,
"loss": 3.5099,
"step": 7800
},
{
"epoch": 2.112510766580534,
"grad_norm": 0.43572768568992615,
"learning_rate": 0.0005686993402450518,
"loss": 3.5095,
"step": 7850
},
{
"epoch": 2.125968992248062,
"grad_norm": 0.3603117763996124,
"learning_rate": 0.000568497374444594,
"loss": 3.511,
"step": 7900
},
{
"epoch": 2.13942721791559,
"grad_norm": 0.3712218105792999,
"learning_rate": 0.0005682954086441362,
"loss": 3.5121,
"step": 7950
},
{
"epoch": 2.152885443583118,
"grad_norm": 0.3927130103111267,
"learning_rate": 0.0005680934428436784,
"loss": 3.5053,
"step": 8000
},
{
"epoch": 2.152885443583118,
"eval_accuracy": 0.3639965180855729,
"eval_loss": 3.539733648300171,
"eval_runtime": 53.0444,
"eval_samples_per_second": 339.565,
"eval_steps_per_second": 21.228,
"step": 8000
},
{
"epoch": 2.166343669250646,
"grad_norm": 0.39150112867355347,
"learning_rate": 0.0005678914770432206,
"loss": 3.502,
"step": 8050
},
{
"epoch": 2.179801894918174,
"grad_norm": 0.4043424129486084,
"learning_rate": 0.0005676895112427629,
"loss": 3.5005,
"step": 8100
},
{
"epoch": 2.193260120585702,
"grad_norm": 0.38725680112838745,
"learning_rate": 0.000567487545442305,
"loss": 3.5122,
"step": 8150
},
{
"epoch": 2.20671834625323,
"grad_norm": 0.36324837803840637,
"learning_rate": 0.0005672855796418473,
"loss": 3.504,
"step": 8200
},
{
"epoch": 2.220176571920758,
"grad_norm": 0.35534292459487915,
"learning_rate": 0.0005670836138413894,
"loss": 3.508,
"step": 8250
},
{
"epoch": 2.233634797588286,
"grad_norm": 0.4022197425365448,
"learning_rate": 0.0005668816480409317,
"loss": 3.5084,
"step": 8300
},
{
"epoch": 2.2470930232558137,
"grad_norm": 0.35888510942459106,
"learning_rate": 0.0005666796822404739,
"loss": 3.5138,
"step": 8350
},
{
"epoch": 2.260551248923342,
"grad_norm": 0.3790094554424286,
"learning_rate": 0.0005664777164400162,
"loss": 3.5104,
"step": 8400
},
{
"epoch": 2.27400947459087,
"grad_norm": 0.37779271602630615,
"learning_rate": 0.0005662757506395583,
"loss": 3.5032,
"step": 8450
},
{
"epoch": 2.287467700258398,
"grad_norm": 0.3711851239204407,
"learning_rate": 0.0005660737848391006,
"loss": 3.5084,
"step": 8500
},
{
"epoch": 2.300925925925926,
"grad_norm": 0.35237640142440796,
"learning_rate": 0.0005658718190386428,
"loss": 3.5112,
"step": 8550
},
{
"epoch": 2.314384151593454,
"grad_norm": 0.401823490858078,
"learning_rate": 0.000565669853238185,
"loss": 3.5101,
"step": 8600
},
{
"epoch": 2.327842377260982,
"grad_norm": 0.39976271986961365,
"learning_rate": 0.0005654678874377272,
"loss": 3.5018,
"step": 8650
},
{
"epoch": 2.3413006029285097,
"grad_norm": 0.37438446283340454,
"learning_rate": 0.0005652659216372693,
"loss": 3.4992,
"step": 8700
},
{
"epoch": 2.354758828596038,
"grad_norm": 0.37986740469932556,
"learning_rate": 0.0005650639558368116,
"loss": 3.4967,
"step": 8750
},
{
"epoch": 2.3682170542635657,
"grad_norm": 0.3729777932167053,
"learning_rate": 0.0005648619900363538,
"loss": 3.4907,
"step": 8800
},
{
"epoch": 2.381675279931094,
"grad_norm": 0.4080381989479065,
"learning_rate": 0.000564660024235896,
"loss": 3.5052,
"step": 8850
},
{
"epoch": 2.395133505598622,
"grad_norm": 0.3736269176006317,
"learning_rate": 0.0005644580584354382,
"loss": 3.4969,
"step": 8900
},
{
"epoch": 2.40859173126615,
"grad_norm": 0.39896562695503235,
"learning_rate": 0.0005642560926349804,
"loss": 3.4992,
"step": 8950
},
{
"epoch": 2.422049956933678,
"grad_norm": 0.3957862854003906,
"learning_rate": 0.0005640541268345226,
"loss": 3.4997,
"step": 9000
},
{
"epoch": 2.422049956933678,
"eval_accuracy": 0.3664585644795185,
"eval_loss": 3.513334035873413,
"eval_runtime": 53.2081,
"eval_samples_per_second": 338.52,
"eval_steps_per_second": 21.162,
"step": 9000
},
{
"epoch": 2.4355081826012057,
"grad_norm": 0.4161396920681,
"learning_rate": 0.0005638521610340649,
"loss": 3.4993,
"step": 9050
},
{
"epoch": 2.448966408268734,
"grad_norm": 0.3905985951423645,
"learning_rate": 0.000563650195233607,
"loss": 3.4938,
"step": 9100
},
{
"epoch": 2.4624246339362617,
"grad_norm": 0.3634888529777527,
"learning_rate": 0.0005634482294331493,
"loss": 3.4957,
"step": 9150
},
{
"epoch": 2.47588285960379,
"grad_norm": 0.34948739409446716,
"learning_rate": 0.0005632462636326914,
"loss": 3.5059,
"step": 9200
},
{
"epoch": 2.4893410852713176,
"grad_norm": 0.3566109538078308,
"learning_rate": 0.0005630442978322337,
"loss": 3.5028,
"step": 9250
},
{
"epoch": 2.502799310938846,
"grad_norm": 0.3608540892601013,
"learning_rate": 0.0005628423320317759,
"loss": 3.4995,
"step": 9300
},
{
"epoch": 2.516257536606374,
"grad_norm": 0.3242948055267334,
"learning_rate": 0.0005626403662313182,
"loss": 3.4847,
"step": 9350
},
{
"epoch": 2.5297157622739017,
"grad_norm": 0.40635618567466736,
"learning_rate": 0.0005624384004308603,
"loss": 3.4935,
"step": 9400
},
{
"epoch": 2.54317398794143,
"grad_norm": 0.34435972571372986,
"learning_rate": 0.0005622364346304026,
"loss": 3.4989,
"step": 9450
},
{
"epoch": 2.5566322136089576,
"grad_norm": 0.44505786895751953,
"learning_rate": 0.0005620344688299448,
"loss": 3.4959,
"step": 9500
},
{
"epoch": 2.570090439276486,
"grad_norm": 0.36672013998031616,
"learning_rate": 0.0005618325030294869,
"loss": 3.4928,
"step": 9550
},
{
"epoch": 2.583548664944014,
"grad_norm": 0.37671077251434326,
"learning_rate": 0.0005616305372290292,
"loss": 3.4988,
"step": 9600
},
{
"epoch": 2.5970068906115418,
"grad_norm": 0.37526118755340576,
"learning_rate": 0.0005614285714285713,
"loss": 3.4852,
"step": 9650
},
{
"epoch": 2.6104651162790695,
"grad_norm": 0.4347611963748932,
"learning_rate": 0.0005612266056281136,
"loss": 3.4959,
"step": 9700
},
{
"epoch": 2.6239233419465977,
"grad_norm": 0.34966251254081726,
"learning_rate": 0.0005610246398276558,
"loss": 3.4847,
"step": 9750
},
{
"epoch": 2.637381567614126,
"grad_norm": 0.3655852675437927,
"learning_rate": 0.000560822674027198,
"loss": 3.4847,
"step": 9800
},
{
"epoch": 2.6508397932816536,
"grad_norm": 0.3635706305503845,
"learning_rate": 0.0005606207082267402,
"loss": 3.4832,
"step": 9850
},
{
"epoch": 2.664298018949182,
"grad_norm": 0.38875824213027954,
"learning_rate": 0.0005604187424262824,
"loss": 3.4892,
"step": 9900
},
{
"epoch": 2.6777562446167096,
"grad_norm": 0.35097113251686096,
"learning_rate": 0.0005602167766258246,
"loss": 3.4833,
"step": 9950
},
{
"epoch": 2.6912144702842378,
"grad_norm": 0.38976341485977173,
"learning_rate": 0.0005600148108253669,
"loss": 3.4848,
"step": 10000
},
{
"epoch": 2.6912144702842378,
"eval_accuracy": 0.36907423752723234,
"eval_loss": 3.487663507461548,
"eval_runtime": 53.3765,
"eval_samples_per_second": 337.452,
"eval_steps_per_second": 21.095,
"step": 10000
},
{
"epoch": 2.704672695951766,
"grad_norm": 0.37565913796424866,
"learning_rate": 0.000559812845024909,
"loss": 3.4913,
"step": 10050
},
{
"epoch": 2.7181309216192937,
"grad_norm": 0.37652722001075745,
"learning_rate": 0.0005596108792244513,
"loss": 3.4784,
"step": 10100
},
{
"epoch": 2.7315891472868215,
"grad_norm": 0.33579641580581665,
"learning_rate": 0.0005594089134239935,
"loss": 3.4749,
"step": 10150
},
{
"epoch": 2.7450473729543496,
"grad_norm": 0.34873345494270325,
"learning_rate": 0.0005592069476235358,
"loss": 3.4802,
"step": 10200
},
{
"epoch": 2.758505598621878,
"grad_norm": 0.362138956785202,
"learning_rate": 0.0005590049818230779,
"loss": 3.4774,
"step": 10250
},
{
"epoch": 2.7719638242894056,
"grad_norm": 0.37752765417099,
"learning_rate": 0.0005588030160226202,
"loss": 3.4785,
"step": 10300
},
{
"epoch": 2.7854220499569338,
"grad_norm": 0.3606274724006653,
"learning_rate": 0.0005586010502221623,
"loss": 3.4773,
"step": 10350
},
{
"epoch": 2.7988802756244615,
"grad_norm": 0.3434750437736511,
"learning_rate": 0.0005583990844217045,
"loss": 3.4756,
"step": 10400
},
{
"epoch": 2.8123385012919897,
"grad_norm": 0.33499059081077576,
"learning_rate": 0.0005581971186212468,
"loss": 3.486,
"step": 10450
},
{
"epoch": 2.825796726959518,
"grad_norm": 0.37345173954963684,
"learning_rate": 0.0005579951528207889,
"loss": 3.4833,
"step": 10500
},
{
"epoch": 2.8392549526270456,
"grad_norm": 0.3695079982280731,
"learning_rate": 0.0005577931870203312,
"loss": 3.4728,
"step": 10550
},
{
"epoch": 2.8527131782945734,
"grad_norm": 0.368105947971344,
"learning_rate": 0.0005575912212198733,
"loss": 3.4682,
"step": 10600
},
{
"epoch": 2.8661714039621016,
"grad_norm": 0.3338841497898102,
"learning_rate": 0.0005573892554194156,
"loss": 3.4753,
"step": 10650
},
{
"epoch": 2.8796296296296298,
"grad_norm": 0.36992380023002625,
"learning_rate": 0.0005571872896189578,
"loss": 3.4641,
"step": 10700
},
{
"epoch": 2.8930878552971575,
"grad_norm": 0.3489851951599121,
"learning_rate": 0.0005569853238185,
"loss": 3.4734,
"step": 10750
},
{
"epoch": 2.9065460809646857,
"grad_norm": 0.35536548495292664,
"learning_rate": 0.0005567833580180422,
"loss": 3.4694,
"step": 10800
},
{
"epoch": 2.9200043066322134,
"grad_norm": 0.3439956307411194,
"learning_rate": 0.0005565813922175844,
"loss": 3.4861,
"step": 10850
},
{
"epoch": 2.9334625322997416,
"grad_norm": 0.32711493968963623,
"learning_rate": 0.0005563794264171266,
"loss": 3.4761,
"step": 10900
},
{
"epoch": 2.94692075796727,
"grad_norm": 0.36705949902534485,
"learning_rate": 0.0005561774606166689,
"loss": 3.4692,
"step": 10950
},
{
"epoch": 2.9603789836347976,
"grad_norm": 0.3396705389022827,
"learning_rate": 0.000555975494816211,
"loss": 3.4716,
"step": 11000
},
{
"epoch": 2.9603789836347976,
"eval_accuracy": 0.37102390535033614,
"eval_loss": 3.4675471782684326,
"eval_runtime": 53.1554,
"eval_samples_per_second": 338.856,
"eval_steps_per_second": 21.183,
"step": 11000
},
{
"epoch": 2.9738372093023253,
"grad_norm": 0.4038015902042389,
"learning_rate": 0.0005557735290157534,
"loss": 3.4765,
"step": 11050
},
{
"epoch": 2.9872954349698535,
"grad_norm": 0.3803868889808655,
"learning_rate": 0.0005555715632152955,
"loss": 3.4671,
"step": 11100
},
{
"epoch": 3.000538329026701,
"grad_norm": 0.4042735695838928,
"learning_rate": 0.0005553695974148378,
"loss": 3.467,
"step": 11150
},
{
"epoch": 3.013996554694229,
"grad_norm": 0.35812151432037354,
"learning_rate": 0.0005551676316143799,
"loss": 3.3741,
"step": 11200
},
{
"epoch": 3.027454780361757,
"grad_norm": 0.36608004570007324,
"learning_rate": 0.0005549656658139222,
"loss": 3.3646,
"step": 11250
},
{
"epoch": 3.0409130060292853,
"grad_norm": 0.3438897430896759,
"learning_rate": 0.0005547637000134644,
"loss": 3.3726,
"step": 11300
},
{
"epoch": 3.054371231696813,
"grad_norm": 0.36485397815704346,
"learning_rate": 0.0005545617342130065,
"loss": 3.3717,
"step": 11350
},
{
"epoch": 3.067829457364341,
"grad_norm": 0.371673583984375,
"learning_rate": 0.0005543597684125488,
"loss": 3.3852,
"step": 11400
},
{
"epoch": 3.081287683031869,
"grad_norm": 0.37548917531967163,
"learning_rate": 0.0005541578026120909,
"loss": 3.3954,
"step": 11450
},
{
"epoch": 3.094745908699397,
"grad_norm": 0.3658439815044403,
"learning_rate": 0.0005539558368116332,
"loss": 3.3921,
"step": 11500
},
{
"epoch": 3.108204134366925,
"grad_norm": 0.3572917878627777,
"learning_rate": 0.0005537538710111754,
"loss": 3.3885,
"step": 11550
},
{
"epoch": 3.121662360034453,
"grad_norm": 0.3559921383857727,
"learning_rate": 0.0005535519052107176,
"loss": 3.3961,
"step": 11600
},
{
"epoch": 3.135120585701981,
"grad_norm": 0.34737083315849304,
"learning_rate": 0.0005533499394102598,
"loss": 3.3874,
"step": 11650
},
{
"epoch": 3.148578811369509,
"grad_norm": 0.3417198061943054,
"learning_rate": 0.000553147973609802,
"loss": 3.4002,
"step": 11700
},
{
"epoch": 3.162037037037037,
"grad_norm": 0.34037500619888306,
"learning_rate": 0.0005529460078093442,
"loss": 3.3855,
"step": 11750
},
{
"epoch": 3.175495262704565,
"grad_norm": 0.36747488379478455,
"learning_rate": 0.0005527440420088865,
"loss": 3.3956,
"step": 11800
},
{
"epoch": 3.188953488372093,
"grad_norm": 0.3483152389526367,
"learning_rate": 0.0005525420762084286,
"loss": 3.3962,
"step": 11850
},
{
"epoch": 3.202411714039621,
"grad_norm": 0.3729177415370941,
"learning_rate": 0.0005523401104079709,
"loss": 3.4056,
"step": 11900
},
{
"epoch": 3.215869939707149,
"grad_norm": 0.34760040044784546,
"learning_rate": 0.0005521381446075131,
"loss": 3.4,
"step": 11950
},
{
"epoch": 3.229328165374677,
"grad_norm": 0.37514349818229675,
"learning_rate": 0.0005519361788070554,
"loss": 3.4002,
"step": 12000
},
{
"epoch": 3.229328165374677,
"eval_accuracy": 0.372848086055263,
"eval_loss": 3.457045078277588,
"eval_runtime": 53.5082,
"eval_samples_per_second": 336.622,
"eval_steps_per_second": 21.044,
"step": 12000
},
{
"epoch": 3.242786391042205,
"grad_norm": 0.34385743737220764,
"learning_rate": 0.0005517342130065975,
"loss": 3.3903,
"step": 12050
},
{
"epoch": 3.2562446167097328,
"grad_norm": 0.32674697041511536,
"learning_rate": 0.0005515322472061398,
"loss": 3.4041,
"step": 12100
},
{
"epoch": 3.269702842377261,
"grad_norm": 0.34761109948158264,
"learning_rate": 0.0005513302814056819,
"loss": 3.3983,
"step": 12150
},
{
"epoch": 3.283161068044789,
"grad_norm": 0.36604440212249756,
"learning_rate": 0.0005511283156052242,
"loss": 3.4032,
"step": 12200
},
{
"epoch": 3.296619293712317,
"grad_norm": 0.35515835881233215,
"learning_rate": 0.0005509263498047664,
"loss": 3.4022,
"step": 12250
},
{
"epoch": 3.310077519379845,
"grad_norm": 0.3531981110572815,
"learning_rate": 0.0005507243840043085,
"loss": 3.397,
"step": 12300
},
{
"epoch": 3.323535745047373,
"grad_norm": 0.337921679019928,
"learning_rate": 0.0005505224182038508,
"loss": 3.4047,
"step": 12350
},
{
"epoch": 3.336993970714901,
"grad_norm": 0.3581739664077759,
"learning_rate": 0.0005503204524033929,
"loss": 3.4017,
"step": 12400
},
{
"epoch": 3.3504521963824287,
"grad_norm": 0.36727410554885864,
"learning_rate": 0.0005501184866029352,
"loss": 3.3959,
"step": 12450
},
{
"epoch": 3.363910422049957,
"grad_norm": 0.35325637459754944,
"learning_rate": 0.0005499165208024774,
"loss": 3.3999,
"step": 12500
},
{
"epoch": 3.3773686477174847,
"grad_norm": 0.3525833785533905,
"learning_rate": 0.0005497145550020196,
"loss": 3.4057,
"step": 12550
},
{
"epoch": 3.390826873385013,
"grad_norm": 0.3457193970680237,
"learning_rate": 0.0005495125892015618,
"loss": 3.4155,
"step": 12600
},
{
"epoch": 3.404285099052541,
"grad_norm": 0.37200361490249634,
"learning_rate": 0.000549310623401104,
"loss": 3.4005,
"step": 12650
},
{
"epoch": 3.417743324720069,
"grad_norm": 0.34540146589279175,
"learning_rate": 0.0005491086576006462,
"loss": 3.4055,
"step": 12700
},
{
"epoch": 3.431201550387597,
"grad_norm": 0.35365238785743713,
"learning_rate": 0.0005489066918001885,
"loss": 3.3988,
"step": 12750
},
{
"epoch": 3.4446597760551247,
"grad_norm": 0.35095787048339844,
"learning_rate": 0.0005487047259997306,
"loss": 3.3998,
"step": 12800
},
{
"epoch": 3.458118001722653,
"grad_norm": 0.34939098358154297,
"learning_rate": 0.0005485027601992729,
"loss": 3.3908,
"step": 12850
},
{
"epoch": 3.471576227390181,
"grad_norm": 0.3667967617511749,
"learning_rate": 0.0005483007943988151,
"loss": 3.396,
"step": 12900
},
{
"epoch": 3.485034453057709,
"grad_norm": 0.357086181640625,
"learning_rate": 0.0005480988285983574,
"loss": 3.3958,
"step": 12950
},
{
"epoch": 3.4984926787252366,
"grad_norm": 0.37680062651634216,
"learning_rate": 0.0005478968627978995,
"loss": 3.401,
"step": 13000
},
{
"epoch": 3.4984926787252366,
"eval_accuracy": 0.3740813365127749,
"eval_loss": 3.4411723613739014,
"eval_runtime": 53.1367,
"eval_samples_per_second": 338.975,
"eval_steps_per_second": 21.191,
"step": 13000
},
{
"epoch": 3.511950904392765,
"grad_norm": 0.3376721441745758,
"learning_rate": 0.0005476948969974418,
"loss": 3.4104,
"step": 13050
},
{
"epoch": 3.525409130060293,
"grad_norm": 0.38172057271003723,
"learning_rate": 0.0005474929311969839,
"loss": 3.4016,
"step": 13100
},
{
"epoch": 3.5388673557278207,
"grad_norm": 0.38476553559303284,
"learning_rate": 0.0005472909653965261,
"loss": 3.3922,
"step": 13150
},
{
"epoch": 3.552325581395349,
"grad_norm": 0.3454050123691559,
"learning_rate": 0.0005470889995960684,
"loss": 3.3904,
"step": 13200
},
{
"epoch": 3.5657838070628767,
"grad_norm": 0.33698442578315735,
"learning_rate": 0.0005468870337956105,
"loss": 3.4087,
"step": 13250
},
{
"epoch": 3.579242032730405,
"grad_norm": 0.32757994532585144,
"learning_rate": 0.0005466850679951528,
"loss": 3.3995,
"step": 13300
},
{
"epoch": 3.592700258397933,
"grad_norm": 0.3386076092720032,
"learning_rate": 0.0005464831021946949,
"loss": 3.3935,
"step": 13350
},
{
"epoch": 3.606158484065461,
"grad_norm": 0.33555930852890015,
"learning_rate": 0.0005462811363942372,
"loss": 3.3903,
"step": 13400
},
{
"epoch": 3.6196167097329885,
"grad_norm": 0.333196759223938,
"learning_rate": 0.0005460791705937794,
"loss": 3.4044,
"step": 13450
},
{
"epoch": 3.6330749354005167,
"grad_norm": 0.3701780438423157,
"learning_rate": 0.0005458772047933216,
"loss": 3.3974,
"step": 13500
},
{
"epoch": 3.646533161068045,
"grad_norm": 0.3696497976779938,
"learning_rate": 0.0005456752389928638,
"loss": 3.4092,
"step": 13550
},
{
"epoch": 3.6599913867355727,
"grad_norm": 0.3531237840652466,
"learning_rate": 0.000545473273192406,
"loss": 3.3965,
"step": 13600
},
{
"epoch": 3.673449612403101,
"grad_norm": 0.3437979817390442,
"learning_rate": 0.0005452713073919482,
"loss": 3.4056,
"step": 13650
},
{
"epoch": 3.6869078380706286,
"grad_norm": 0.3295712471008301,
"learning_rate": 0.0005450693415914905,
"loss": 3.4118,
"step": 13700
},
{
"epoch": 3.700366063738157,
"grad_norm": 0.39618441462516785,
"learning_rate": 0.0005448673757910327,
"loss": 3.4115,
"step": 13750
},
{
"epoch": 3.713824289405685,
"grad_norm": 0.37704867124557495,
"learning_rate": 0.0005446654099905749,
"loss": 3.3902,
"step": 13800
},
{
"epoch": 3.7272825150732127,
"grad_norm": 0.3658808171749115,
"learning_rate": 0.0005444634441901171,
"loss": 3.4064,
"step": 13850
},
{
"epoch": 3.7407407407407405,
"grad_norm": 0.401401549577713,
"learning_rate": 0.0005442614783896594,
"loss": 3.3959,
"step": 13900
},
{
"epoch": 3.7541989664082687,
"grad_norm": 0.33702999353408813,
"learning_rate": 0.0005440595125892015,
"loss": 3.3956,
"step": 13950
},
{
"epoch": 3.767657192075797,
"grad_norm": 0.36874791979789734,
"learning_rate": 0.0005438575467887438,
"loss": 3.3903,
"step": 14000
},
{
"epoch": 3.767657192075797,
"eval_accuracy": 0.37602872275191185,
"eval_loss": 3.4243662357330322,
"eval_runtime": 53.3346,
"eval_samples_per_second": 337.717,
"eval_steps_per_second": 21.112,
"step": 14000
},
{
"epoch": 3.7811154177433246,
"grad_norm": 0.3917904198169708,
"learning_rate": 0.0005436555809882859,
"loss": 3.4051,
"step": 14050
},
{
"epoch": 3.794573643410853,
"grad_norm": 0.367305189371109,
"learning_rate": 0.0005434536151878281,
"loss": 3.4014,
"step": 14100
},
{
"epoch": 3.8080318690783805,
"grad_norm": 0.3374054431915283,
"learning_rate": 0.0005432516493873704,
"loss": 3.3996,
"step": 14150
},
{
"epoch": 3.8214900947459087,
"grad_norm": 0.3432117700576782,
"learning_rate": 0.0005430496835869125,
"loss": 3.3921,
"step": 14200
},
{
"epoch": 3.834948320413437,
"grad_norm": 0.36590588092803955,
"learning_rate": 0.0005428477177864548,
"loss": 3.4038,
"step": 14250
},
{
"epoch": 3.8484065460809647,
"grad_norm": 0.4062247574329376,
"learning_rate": 0.0005426457519859969,
"loss": 3.4078,
"step": 14300
},
{
"epoch": 3.8618647717484924,
"grad_norm": 0.31158193945884705,
"learning_rate": 0.0005424437861855392,
"loss": 3.3981,
"step": 14350
},
{
"epoch": 3.8753229974160206,
"grad_norm": 0.3665839433670044,
"learning_rate": 0.0005422418203850814,
"loss": 3.3995,
"step": 14400
},
{
"epoch": 3.888781223083549,
"grad_norm": 0.34909674525260925,
"learning_rate": 0.0005420398545846236,
"loss": 3.3967,
"step": 14450
},
{
"epoch": 3.9022394487510765,
"grad_norm": 0.34972283244132996,
"learning_rate": 0.0005418378887841658,
"loss": 3.4018,
"step": 14500
},
{
"epoch": 3.9156976744186047,
"grad_norm": 0.36741903424263,
"learning_rate": 0.000541635922983708,
"loss": 3.3922,
"step": 14550
},
{
"epoch": 3.9291559000861325,
"grad_norm": 0.34265077114105225,
"learning_rate": 0.0005414339571832503,
"loss": 3.3984,
"step": 14600
},
{
"epoch": 3.9426141257536607,
"grad_norm": 0.35221385955810547,
"learning_rate": 0.0005412319913827925,
"loss": 3.3915,
"step": 14650
},
{
"epoch": 3.956072351421189,
"grad_norm": 0.3693876266479492,
"learning_rate": 0.0005410300255823347,
"loss": 3.4048,
"step": 14700
},
{
"epoch": 3.9695305770887166,
"grad_norm": 0.36447668075561523,
"learning_rate": 0.0005408280597818769,
"loss": 3.4002,
"step": 14750
},
{
"epoch": 3.9829888027562443,
"grad_norm": 0.37166357040405273,
"learning_rate": 0.0005406260939814191,
"loss": 3.3946,
"step": 14800
},
{
"epoch": 3.9964470284237725,
"grad_norm": 0.3421415090560913,
"learning_rate": 0.0005404241281809614,
"loss": 3.391,
"step": 14850
},
{
"epoch": 4.00968992248062,
"grad_norm": 0.3298637866973877,
"learning_rate": 0.0005402221623805035,
"loss": 3.3309,
"step": 14900
},
{
"epoch": 4.023148148148148,
"grad_norm": 0.35694819688796997,
"learning_rate": 0.0005400201965800457,
"loss": 3.2933,
"step": 14950
},
{
"epoch": 4.036606373815676,
"grad_norm": 0.3794950842857361,
"learning_rate": 0.000539818230779588,
"loss": 3.2948,
"step": 15000
},
{
"epoch": 4.036606373815676,
"eval_accuracy": 0.37712822893022396,
"eval_loss": 3.417656183242798,
"eval_runtime": 53.6437,
"eval_samples_per_second": 335.771,
"eval_steps_per_second": 20.99,
"step": 15000
},
{
"epoch": 4.050064599483204,
"grad_norm": 0.366841584444046,
"learning_rate": 0.0005396162649791301,
"loss": 3.312,
"step": 15050
},
{
"epoch": 4.0635228251507325,
"grad_norm": 0.3495643138885498,
"learning_rate": 0.0005394142991786724,
"loss": 3.3137,
"step": 15100
},
{
"epoch": 4.07698105081826,
"grad_norm": 0.3431321382522583,
"learning_rate": 0.0005392123333782145,
"loss": 3.3048,
"step": 15150
},
{
"epoch": 4.090439276485788,
"grad_norm": 0.36576730012893677,
"learning_rate": 0.0005390103675777568,
"loss": 3.313,
"step": 15200
},
{
"epoch": 4.103897502153316,
"grad_norm": 0.36886483430862427,
"learning_rate": 0.000538808401777299,
"loss": 3.3163,
"step": 15250
},
{
"epoch": 4.117355727820844,
"grad_norm": 0.35335373878479004,
"learning_rate": 0.0005386064359768412,
"loss": 3.3143,
"step": 15300
},
{
"epoch": 4.1308139534883725,
"grad_norm": 0.35200396180152893,
"learning_rate": 0.0005384044701763834,
"loss": 3.3102,
"step": 15350
},
{
"epoch": 4.1442721791559,
"grad_norm": 0.331868052482605,
"learning_rate": 0.0005382025043759256,
"loss": 3.3176,
"step": 15400
},
{
"epoch": 4.157730404823428,
"grad_norm": 0.3521571159362793,
"learning_rate": 0.0005380005385754678,
"loss": 3.3228,
"step": 15450
},
{
"epoch": 4.171188630490956,
"grad_norm": 0.3609760105609894,
"learning_rate": 0.0005377985727750101,
"loss": 3.2974,
"step": 15500
},
{
"epoch": 4.184646856158484,
"grad_norm": 0.36947956681251526,
"learning_rate": 0.0005375966069745523,
"loss": 3.3197,
"step": 15550
},
{
"epoch": 4.198105081826012,
"grad_norm": 0.3476182222366333,
"learning_rate": 0.0005373946411740945,
"loss": 3.319,
"step": 15600
},
{
"epoch": 4.21156330749354,
"grad_norm": 0.3870675265789032,
"learning_rate": 0.0005371926753736367,
"loss": 3.3361,
"step": 15650
},
{
"epoch": 4.225021533161068,
"grad_norm": 0.33378762006759644,
"learning_rate": 0.000536990709573179,
"loss": 3.3219,
"step": 15700
},
{
"epoch": 4.238479758828596,
"grad_norm": 0.3631652593612671,
"learning_rate": 0.0005367887437727211,
"loss": 3.3383,
"step": 15750
},
{
"epoch": 4.251937984496124,
"grad_norm": 0.34275636076927185,
"learning_rate": 0.0005365867779722634,
"loss": 3.3245,
"step": 15800
},
{
"epoch": 4.265396210163652,
"grad_norm": 0.3362019956111908,
"learning_rate": 0.0005363848121718055,
"loss": 3.3262,
"step": 15850
},
{
"epoch": 4.27885443583118,
"grad_norm": 0.32696837186813354,
"learning_rate": 0.0005361828463713477,
"loss": 3.3358,
"step": 15900
},
{
"epoch": 4.292312661498708,
"grad_norm": 0.3329922556877136,
"learning_rate": 0.00053598088057089,
"loss": 3.3274,
"step": 15950
},
{
"epoch": 4.305770887166236,
"grad_norm": 0.3461296260356903,
"learning_rate": 0.0005357789147704321,
"loss": 3.332,
"step": 16000
},
{
"epoch": 4.305770887166236,
"eval_accuracy": 0.37815222554391875,
"eval_loss": 3.406710147857666,
"eval_runtime": 53.6423,
"eval_samples_per_second": 335.78,
"eval_steps_per_second": 20.991,
"step": 16000
},
{
"epoch": 4.319229112833764,
"grad_norm": 0.36084792017936707,
"learning_rate": 0.0005355769489699744,
"loss": 3.3323,
"step": 16050
},
{
"epoch": 4.332687338501292,
"grad_norm": 0.39574679732322693,
"learning_rate": 0.0005353749831695165,
"loss": 3.334,
"step": 16100
},
{
"epoch": 4.34614556416882,
"grad_norm": 0.3344963788986206,
"learning_rate": 0.0005351730173690588,
"loss": 3.3347,
"step": 16150
},
{
"epoch": 4.359603789836348,
"grad_norm": 0.33946841955184937,
"learning_rate": 0.000534971051568601,
"loss": 3.3294,
"step": 16200
},
{
"epoch": 4.373062015503876,
"grad_norm": 0.35259753465652466,
"learning_rate": 0.0005347690857681432,
"loss": 3.3383,
"step": 16250
},
{
"epoch": 4.386520241171404,
"grad_norm": 0.3345356583595276,
"learning_rate": 0.0005345671199676854,
"loss": 3.3435,
"step": 16300
},
{
"epoch": 4.399978466838932,
"grad_norm": 0.33119991421699524,
"learning_rate": 0.0005343651541672276,
"loss": 3.337,
"step": 16350
},
{
"epoch": 4.41343669250646,
"grad_norm": 0.33489561080932617,
"learning_rate": 0.0005341631883667699,
"loss": 3.3357,
"step": 16400
},
{
"epoch": 4.426894918173988,
"grad_norm": 0.36713531613349915,
"learning_rate": 0.0005339612225663121,
"loss": 3.3417,
"step": 16450
},
{
"epoch": 4.440353143841516,
"grad_norm": 0.3532050549983978,
"learning_rate": 0.0005337592567658543,
"loss": 3.3358,
"step": 16500
},
{
"epoch": 4.453811369509044,
"grad_norm": 0.3615681529045105,
"learning_rate": 0.0005335572909653965,
"loss": 3.3381,
"step": 16550
},
{
"epoch": 4.467269595176572,
"grad_norm": 0.3489953279495239,
"learning_rate": 0.0005333553251649387,
"loss": 3.3482,
"step": 16600
},
{
"epoch": 4.4807278208441,
"grad_norm": 0.354987233877182,
"learning_rate": 0.000533153359364481,
"loss": 3.3319,
"step": 16650
},
{
"epoch": 4.4941860465116275,
"grad_norm": 0.35782018303871155,
"learning_rate": 0.0005329513935640231,
"loss": 3.3424,
"step": 16700
},
{
"epoch": 4.507644272179156,
"grad_norm": 0.360448956489563,
"learning_rate": 0.0005327494277635654,
"loss": 3.3438,
"step": 16750
},
{
"epoch": 4.521102497846684,
"grad_norm": 0.38995808362960815,
"learning_rate": 0.0005325474619631075,
"loss": 3.3502,
"step": 16800
},
{
"epoch": 4.534560723514212,
"grad_norm": 0.3384845554828644,
"learning_rate": 0.0005323454961626497,
"loss": 3.3371,
"step": 16850
},
{
"epoch": 4.54801894918174,
"grad_norm": 0.3251652717590332,
"learning_rate": 0.000532143530362192,
"loss": 3.3356,
"step": 16900
},
{
"epoch": 4.5614771748492675,
"grad_norm": 0.34360337257385254,
"learning_rate": 0.0005319415645617341,
"loss": 3.3334,
"step": 16950
},
{
"epoch": 4.574935400516796,
"grad_norm": 0.35140931606292725,
"learning_rate": 0.0005317395987612764,
"loss": 3.3342,
"step": 17000
},
{
"epoch": 4.574935400516796,
"eval_accuracy": 0.37956517790053423,
"eval_loss": 3.3962960243225098,
"eval_runtime": 53.7518,
"eval_samples_per_second": 335.096,
"eval_steps_per_second": 20.948,
"step": 17000
},
{
"epoch": 4.588393626184324,
"grad_norm": 0.334964781999588,
"learning_rate": 0.0005315376329608185,
"loss": 3.3331,
"step": 17050
},
{
"epoch": 4.601851851851852,
"grad_norm": 0.33990418910980225,
"learning_rate": 0.0005313356671603608,
"loss": 3.3411,
"step": 17100
},
{
"epoch": 4.61531007751938,
"grad_norm": 0.3458724319934845,
"learning_rate": 0.000531133701359903,
"loss": 3.3441,
"step": 17150
},
{
"epoch": 4.628768303186908,
"grad_norm": 0.35447219014167786,
"learning_rate": 0.0005309317355594452,
"loss": 3.3376,
"step": 17200
},
{
"epoch": 4.642226528854436,
"grad_norm": 0.34828895330429077,
"learning_rate": 0.0005307297697589874,
"loss": 3.3317,
"step": 17250
},
{
"epoch": 4.655684754521964,
"grad_norm": 0.33856168389320374,
"learning_rate": 0.0005305278039585297,
"loss": 3.3392,
"step": 17300
},
{
"epoch": 4.669142980189492,
"grad_norm": 0.3794487416744232,
"learning_rate": 0.0005303258381580719,
"loss": 3.3386,
"step": 17350
},
{
"epoch": 4.682601205857019,
"grad_norm": 0.3594615161418915,
"learning_rate": 0.0005301238723576141,
"loss": 3.3562,
"step": 17400
},
{
"epoch": 4.696059431524548,
"grad_norm": 0.32063180208206177,
"learning_rate": 0.0005299219065571563,
"loss": 3.3455,
"step": 17450
},
{
"epoch": 4.709517657192076,
"grad_norm": 0.34606829285621643,
"learning_rate": 0.0005297199407566985,
"loss": 3.3328,
"step": 17500
},
{
"epoch": 4.722975882859604,
"grad_norm": 0.3413867652416229,
"learning_rate": 0.0005295179749562407,
"loss": 3.3536,
"step": 17550
},
{
"epoch": 4.736434108527131,
"grad_norm": 0.34495651721954346,
"learning_rate": 0.000529316009155783,
"loss": 3.3371,
"step": 17600
},
{
"epoch": 4.7498923341946595,
"grad_norm": 0.3626272678375244,
"learning_rate": 0.0005291140433553251,
"loss": 3.3512,
"step": 17650
},
{
"epoch": 4.763350559862188,
"grad_norm": 0.3632408380508423,
"learning_rate": 0.0005289120775548673,
"loss": 3.3337,
"step": 17700
},
{
"epoch": 4.776808785529716,
"grad_norm": 0.3298455774784088,
"learning_rate": 0.0005287101117544095,
"loss": 3.3392,
"step": 17750
},
{
"epoch": 4.790267011197244,
"grad_norm": 0.3547931909561157,
"learning_rate": 0.0005285081459539517,
"loss": 3.3377,
"step": 17800
},
{
"epoch": 4.803725236864771,
"grad_norm": 0.3601723611354828,
"learning_rate": 0.000528306180153494,
"loss": 3.3349,
"step": 17850
},
{
"epoch": 4.8171834625323,
"grad_norm": 0.3788248598575592,
"learning_rate": 0.0005281042143530361,
"loss": 3.3452,
"step": 17900
},
{
"epoch": 4.830641688199828,
"grad_norm": 0.36663538217544556,
"learning_rate": 0.0005279022485525784,
"loss": 3.3385,
"step": 17950
},
{
"epoch": 4.844099913867356,
"grad_norm": 0.3512565791606903,
"learning_rate": 0.0005277002827521205,
"loss": 3.3418,
"step": 18000
},
{
"epoch": 4.844099913867356,
"eval_accuracy": 0.380534307852169,
"eval_loss": 3.3869707584381104,
"eval_runtime": 53.4631,
"eval_samples_per_second": 336.905,
"eval_steps_per_second": 21.061,
"step": 18000
},
{
"epoch": 4.857558139534884,
"grad_norm": 0.38106250762939453,
"learning_rate": 0.0005274983169516628,
"loss": 3.3281,
"step": 18050
},
{
"epoch": 4.871016365202411,
"grad_norm": 0.32285740971565247,
"learning_rate": 0.000527296351151205,
"loss": 3.3402,
"step": 18100
},
{
"epoch": 4.88447459086994,
"grad_norm": 0.33819901943206787,
"learning_rate": 0.0005270943853507472,
"loss": 3.3322,
"step": 18150
},
{
"epoch": 4.897932816537468,
"grad_norm": 0.3418741822242737,
"learning_rate": 0.0005268924195502894,
"loss": 3.3459,
"step": 18200
},
{
"epoch": 4.911391042204996,
"grad_norm": 0.35104596614837646,
"learning_rate": 0.0005266904537498317,
"loss": 3.3498,
"step": 18250
},
{
"epoch": 4.924849267872523,
"grad_norm": 0.3623558282852173,
"learning_rate": 0.0005264884879493739,
"loss": 3.3433,
"step": 18300
},
{
"epoch": 4.9383074935400515,
"grad_norm": 0.34203389286994934,
"learning_rate": 0.0005262865221489161,
"loss": 3.3313,
"step": 18350
},
{
"epoch": 4.95176571920758,
"grad_norm": 0.34537750482559204,
"learning_rate": 0.0005260845563484583,
"loss": 3.328,
"step": 18400
},
{
"epoch": 4.965223944875108,
"grad_norm": 0.3605607748031616,
"learning_rate": 0.0005258825905480005,
"loss": 3.3394,
"step": 18450
},
{
"epoch": 4.978682170542635,
"grad_norm": 0.346736878156662,
"learning_rate": 0.0005256806247475427,
"loss": 3.3532,
"step": 18500
},
{
"epoch": 4.992140396210163,
"grad_norm": 0.33791518211364746,
"learning_rate": 0.000525478658947085,
"loss": 3.3278,
"step": 18550
},
{
"epoch": 5.0053832902670115,
"grad_norm": 0.33499035239219666,
"learning_rate": 0.0005252766931466271,
"loss": 3.3047,
"step": 18600
},
{
"epoch": 5.018841515934539,
"grad_norm": 0.38031017780303955,
"learning_rate": 0.0005250747273461693,
"loss": 3.2399,
"step": 18650
},
{
"epoch": 5.032299741602067,
"grad_norm": 0.35386326909065247,
"learning_rate": 0.0005248727615457115,
"loss": 3.2533,
"step": 18700
},
{
"epoch": 5.045757967269595,
"grad_norm": 0.364179402589798,
"learning_rate": 0.0005246707957452537,
"loss": 3.2444,
"step": 18750
},
{
"epoch": 5.059216192937123,
"grad_norm": 0.36041221022605896,
"learning_rate": 0.000524468829944796,
"loss": 3.2541,
"step": 18800
},
{
"epoch": 5.0726744186046515,
"grad_norm": 0.3468455672264099,
"learning_rate": 0.0005242668641443381,
"loss": 3.2394,
"step": 18850
},
{
"epoch": 5.086132644272179,
"grad_norm": 0.34421736001968384,
"learning_rate": 0.0005240648983438804,
"loss": 3.2563,
"step": 18900
},
{
"epoch": 5.099590869939707,
"grad_norm": 0.36943989992141724,
"learning_rate": 0.0005238629325434225,
"loss": 3.2654,
"step": 18950
},
{
"epoch": 5.113049095607235,
"grad_norm": 0.3361397087574005,
"learning_rate": 0.0005236609667429648,
"loss": 3.263,
"step": 19000
},
{
"epoch": 5.113049095607235,
"eval_accuracy": 0.38084492921222773,
"eval_loss": 3.3863155841827393,
"eval_runtime": 53.1852,
"eval_samples_per_second": 338.666,
"eval_steps_per_second": 21.171,
"step": 19000
},
{
"epoch": 5.126507321274763,
"grad_norm": 0.3635759651660919,
"learning_rate": 0.000523459000942507,
"loss": 3.2646,
"step": 19050
},
{
"epoch": 5.139965546942291,
"grad_norm": 0.3344425559043884,
"learning_rate": 0.0005232570351420493,
"loss": 3.2643,
"step": 19100
},
{
"epoch": 5.153423772609819,
"grad_norm": 0.3406955599784851,
"learning_rate": 0.0005230550693415915,
"loss": 3.2682,
"step": 19150
},
{
"epoch": 5.166881998277347,
"grad_norm": 0.35041916370391846,
"learning_rate": 0.0005228531035411337,
"loss": 3.2672,
"step": 19200
},
{
"epoch": 5.180340223944875,
"grad_norm": 0.35396963357925415,
"learning_rate": 0.0005226511377406759,
"loss": 3.2665,
"step": 19250
},
{
"epoch": 5.1937984496124034,
"grad_norm": 0.352710098028183,
"learning_rate": 0.0005224491719402181,
"loss": 3.2818,
"step": 19300
},
{
"epoch": 5.207256675279931,
"grad_norm": 0.37922459840774536,
"learning_rate": 0.0005222472061397603,
"loss": 3.2652,
"step": 19350
},
{
"epoch": 5.220714900947459,
"grad_norm": 0.37260207533836365,
"learning_rate": 0.0005220452403393026,
"loss": 3.2688,
"step": 19400
},
{
"epoch": 5.234173126614987,
"grad_norm": 0.33815115690231323,
"learning_rate": 0.0005218432745388447,
"loss": 3.2846,
"step": 19450
},
{
"epoch": 5.247631352282515,
"grad_norm": 0.3498179316520691,
"learning_rate": 0.000521641308738387,
"loss": 3.2794,
"step": 19500
},
{
"epoch": 5.2610895779500435,
"grad_norm": 0.3579239249229431,
"learning_rate": 0.0005214393429379291,
"loss": 3.2748,
"step": 19550
},
{
"epoch": 5.274547803617571,
"grad_norm": 0.3630642890930176,
"learning_rate": 0.0005212373771374713,
"loss": 3.2801,
"step": 19600
},
{
"epoch": 5.288006029285099,
"grad_norm": 0.33767256140708923,
"learning_rate": 0.0005210354113370136,
"loss": 3.2743,
"step": 19650
},
{
"epoch": 5.301464254952627,
"grad_norm": 0.339458703994751,
"learning_rate": 0.0005208334455365557,
"loss": 3.2685,
"step": 19700
},
{
"epoch": 5.314922480620155,
"grad_norm": 0.3411153256893158,
"learning_rate": 0.000520631479736098,
"loss": 3.274,
"step": 19750
},
{
"epoch": 5.328380706287683,
"grad_norm": 0.3428064286708832,
"learning_rate": 0.0005204295139356401,
"loss": 3.2723,
"step": 19800
},
{
"epoch": 5.341838931955211,
"grad_norm": 0.3319042921066284,
"learning_rate": 0.0005202275481351824,
"loss": 3.2836,
"step": 19850
},
{
"epoch": 5.355297157622739,
"grad_norm": 0.34343960881233215,
"learning_rate": 0.0005200255823347246,
"loss": 3.2763,
"step": 19900
},
{
"epoch": 5.368755383290267,
"grad_norm": 0.3530113101005554,
"learning_rate": 0.0005198236165342669,
"loss": 3.2781,
"step": 19950
},
{
"epoch": 5.3822136089577945,
"grad_norm": 0.35422250628471375,
"learning_rate": 0.000519621650733809,
"loss": 3.2798,
"step": 20000
},
{
"epoch": 5.3822136089577945,
"eval_accuracy": 0.3819953907657995,
"eval_loss": 3.3763461112976074,
"eval_runtime": 53.1398,
"eval_samples_per_second": 338.955,
"eval_steps_per_second": 21.189,
"step": 20000
},
{
"epoch": 5.395671834625323,
"grad_norm": 0.32547879219055176,
"learning_rate": 0.0005194196849333513,
"loss": 3.2788,
"step": 20050
},
{
"epoch": 5.409130060292851,
"grad_norm": 0.3453685939311981,
"learning_rate": 0.0005192177191328935,
"loss": 3.2891,
"step": 20100
},
{
"epoch": 5.422588285960379,
"grad_norm": 0.35231003165245056,
"learning_rate": 0.0005190157533324357,
"loss": 3.2897,
"step": 20150
},
{
"epoch": 5.436046511627907,
"grad_norm": 0.32846277952194214,
"learning_rate": 0.0005188137875319779,
"loss": 3.2868,
"step": 20200
},
{
"epoch": 5.449504737295435,
"grad_norm": 0.32960057258605957,
"learning_rate": 0.00051861182173152,
"loss": 3.2865,
"step": 20250
},
{
"epoch": 5.462962962962963,
"grad_norm": 0.3438560664653778,
"learning_rate": 0.0005184098559310623,
"loss": 3.2808,
"step": 20300
},
{
"epoch": 5.476421188630491,
"grad_norm": 0.33576279878616333,
"learning_rate": 0.0005182078901306046,
"loss": 3.2908,
"step": 20350
},
{
"epoch": 5.489879414298019,
"grad_norm": 0.3566272556781769,
"learning_rate": 0.0005180059243301467,
"loss": 3.2877,
"step": 20400
},
{
"epoch": 5.503337639965547,
"grad_norm": 0.3287167549133301,
"learning_rate": 0.000517803958529689,
"loss": 3.2831,
"step": 20450
},
{
"epoch": 5.516795865633075,
"grad_norm": 0.3539256155490875,
"learning_rate": 0.0005176019927292311,
"loss": 3.2837,
"step": 20500
},
{
"epoch": 5.530254091300603,
"grad_norm": 0.3405372202396393,
"learning_rate": 0.0005174000269287733,
"loss": 3.2892,
"step": 20550
},
{
"epoch": 5.543712316968131,
"grad_norm": 0.3327963948249817,
"learning_rate": 0.0005171980611283156,
"loss": 3.2912,
"step": 20600
},
{
"epoch": 5.557170542635659,
"grad_norm": 0.36150923371315,
"learning_rate": 0.0005169960953278577,
"loss": 3.296,
"step": 20650
},
{
"epoch": 5.5706287683031865,
"grad_norm": 0.34517911076545715,
"learning_rate": 0.0005167941295274,
"loss": 3.2856,
"step": 20700
},
{
"epoch": 5.584086993970715,
"grad_norm": 0.3302103579044342,
"learning_rate": 0.0005165921637269421,
"loss": 3.2814,
"step": 20750
},
{
"epoch": 5.597545219638243,
"grad_norm": 0.3642379641532898,
"learning_rate": 0.0005163901979264844,
"loss": 3.286,
"step": 20800
},
{
"epoch": 5.611003445305771,
"grad_norm": 0.3356991410255432,
"learning_rate": 0.0005161882321260266,
"loss": 3.2943,
"step": 20850
},
{
"epoch": 5.624461670973298,
"grad_norm": 0.34540364146232605,
"learning_rate": 0.0005159862663255689,
"loss": 3.2879,
"step": 20900
},
{
"epoch": 5.637919896640827,
"grad_norm": 0.3615170121192932,
"learning_rate": 0.000515784300525111,
"loss": 3.3059,
"step": 20950
},
{
"epoch": 5.651378122308355,
"grad_norm": 0.36911121010780334,
"learning_rate": 0.0005155823347246533,
"loss": 3.2997,
"step": 21000
},
{
"epoch": 5.651378122308355,
"eval_accuracy": 0.38274592324403867,
"eval_loss": 3.367357015609741,
"eval_runtime": 53.3615,
"eval_samples_per_second": 337.547,
"eval_steps_per_second": 21.101,
"step": 21000
},
{
"epoch": 5.664836347975883,
"grad_norm": 0.3656895160675049,
"learning_rate": 0.0005153803689241955,
"loss": 3.3019,
"step": 21050
},
{
"epoch": 5.678294573643411,
"grad_norm": 0.349932998418808,
"learning_rate": 0.0005151784031237377,
"loss": 3.2888,
"step": 21100
},
{
"epoch": 5.6917527993109385,
"grad_norm": 0.329196035861969,
"learning_rate": 0.0005149764373232799,
"loss": 3.2977,
"step": 21150
},
{
"epoch": 5.705211024978467,
"grad_norm": 0.3490157425403595,
"learning_rate": 0.000514774471522822,
"loss": 3.2962,
"step": 21200
},
{
"epoch": 5.718669250645995,
"grad_norm": 0.4173871874809265,
"learning_rate": 0.0005145725057223643,
"loss": 3.2923,
"step": 21250
},
{
"epoch": 5.732127476313523,
"grad_norm": 0.3462425172328949,
"learning_rate": 0.0005143705399219066,
"loss": 3.2846,
"step": 21300
},
{
"epoch": 5.745585701981051,
"grad_norm": 0.3564380407333374,
"learning_rate": 0.0005141685741214487,
"loss": 3.3123,
"step": 21350
},
{
"epoch": 5.7590439276485785,
"grad_norm": 0.3207317590713501,
"learning_rate": 0.0005139666083209909,
"loss": 3.3018,
"step": 21400
},
{
"epoch": 5.772502153316107,
"grad_norm": 0.32770583033561707,
"learning_rate": 0.0005137646425205331,
"loss": 3.2902,
"step": 21450
},
{
"epoch": 5.785960378983635,
"grad_norm": 0.32427579164505005,
"learning_rate": 0.0005135626767200753,
"loss": 3.3019,
"step": 21500
},
{
"epoch": 5.799418604651163,
"grad_norm": 0.3444927930831909,
"learning_rate": 0.0005133607109196176,
"loss": 3.2984,
"step": 21550
},
{
"epoch": 5.81287683031869,
"grad_norm": 0.3426213264465332,
"learning_rate": 0.0005131587451191597,
"loss": 3.3149,
"step": 21600
},
{
"epoch": 5.826335055986219,
"grad_norm": 0.3312057554721832,
"learning_rate": 0.000512956779318702,
"loss": 3.3019,
"step": 21650
},
{
"epoch": 5.839793281653747,
"grad_norm": 0.35812973976135254,
"learning_rate": 0.0005127548135182441,
"loss": 3.3002,
"step": 21700
},
{
"epoch": 5.853251507321275,
"grad_norm": 0.3365882635116577,
"learning_rate": 0.0005125528477177865,
"loss": 3.2915,
"step": 21750
},
{
"epoch": 5.866709732988802,
"grad_norm": 0.3192456066608429,
"learning_rate": 0.0005123508819173286,
"loss": 3.2924,
"step": 21800
},
{
"epoch": 5.8801679586563305,
"grad_norm": 0.3553847074508667,
"learning_rate": 0.0005121489161168709,
"loss": 3.2976,
"step": 21850
},
{
"epoch": 5.893626184323859,
"grad_norm": 0.33080434799194336,
"learning_rate": 0.000511946950316413,
"loss": 3.2883,
"step": 21900
},
{
"epoch": 5.907084409991387,
"grad_norm": 0.3338245451450348,
"learning_rate": 0.0005117449845159553,
"loss": 3.2925,
"step": 21950
},
{
"epoch": 5.920542635658915,
"grad_norm": 0.3583246171474457,
"learning_rate": 0.0005115430187154975,
"loss": 3.3,
"step": 22000
},
{
"epoch": 5.920542635658915,
"eval_accuracy": 0.3836738760374145,
"eval_loss": 3.3571391105651855,
"eval_runtime": 53.3686,
"eval_samples_per_second": 337.502,
"eval_steps_per_second": 21.099,
"step": 22000
},
{
"epoch": 5.934000861326442,
"grad_norm": 0.3355707824230194,
"learning_rate": 0.0005113410529150397,
"loss": 3.3096,
"step": 22050
},
{
"epoch": 5.9474590869939705,
"grad_norm": 0.38139769434928894,
"learning_rate": 0.0005111390871145819,
"loss": 3.3018,
"step": 22100
},
{
"epoch": 5.960917312661499,
"grad_norm": 0.32805705070495605,
"learning_rate": 0.000510937121314124,
"loss": 3.3019,
"step": 22150
},
{
"epoch": 5.974375538329027,
"grad_norm": 0.3372587561607361,
"learning_rate": 0.0005107351555136663,
"loss": 3.3008,
"step": 22200
},
{
"epoch": 5.987833763996555,
"grad_norm": 0.3290460407733917,
"learning_rate": 0.0005105331897132085,
"loss": 3.2948,
"step": 22250
},
{
"epoch": 6.001076658053402,
"grad_norm": 0.3840121328830719,
"learning_rate": 0.0005103312239127507,
"loss": 3.295,
"step": 22300
},
{
"epoch": 6.0145348837209305,
"grad_norm": 0.33211490511894226,
"learning_rate": 0.0005101292581122929,
"loss": 3.1891,
"step": 22350
},
{
"epoch": 6.027993109388458,
"grad_norm": 0.3395225405693054,
"learning_rate": 0.0005099272923118351,
"loss": 3.2021,
"step": 22400
},
{
"epoch": 6.041451335055986,
"grad_norm": 0.3511013388633728,
"learning_rate": 0.0005097253265113773,
"loss": 3.2057,
"step": 22450
},
{
"epoch": 6.054909560723514,
"grad_norm": 0.365784615278244,
"learning_rate": 0.0005095233607109196,
"loss": 3.2036,
"step": 22500
},
{
"epoch": 6.068367786391042,
"grad_norm": 0.34816962480545044,
"learning_rate": 0.0005093213949104617,
"loss": 3.2033,
"step": 22550
},
{
"epoch": 6.0818260120585705,
"grad_norm": 0.3428509831428528,
"learning_rate": 0.000509119429110004,
"loss": 3.2188,
"step": 22600
},
{
"epoch": 6.095284237726098,
"grad_norm": 0.3539871871471405,
"learning_rate": 0.0005089174633095462,
"loss": 3.2135,
"step": 22650
},
{
"epoch": 6.108742463393626,
"grad_norm": 0.37805479764938354,
"learning_rate": 0.0005087154975090885,
"loss": 3.2219,
"step": 22700
},
{
"epoch": 6.122200689061154,
"grad_norm": 0.40196943283081055,
"learning_rate": 0.0005085135317086306,
"loss": 3.2134,
"step": 22750
},
{
"epoch": 6.135658914728682,
"grad_norm": 0.3432955741882324,
"learning_rate": 0.0005083115659081729,
"loss": 3.2254,
"step": 22800
},
{
"epoch": 6.149117140396211,
"grad_norm": 0.4084476828575134,
"learning_rate": 0.0005081096001077151,
"loss": 3.2232,
"step": 22850
},
{
"epoch": 6.162575366063738,
"grad_norm": 0.364326536655426,
"learning_rate": 0.0005079076343072573,
"loss": 3.2344,
"step": 22900
},
{
"epoch": 6.176033591731266,
"grad_norm": 0.34070107340812683,
"learning_rate": 0.0005077056685067995,
"loss": 3.229,
"step": 22950
},
{
"epoch": 6.189491817398794,
"grad_norm": 0.34302592277526855,
"learning_rate": 0.0005075037027063417,
"loss": 3.2446,
"step": 23000
},
{
"epoch": 6.189491817398794,
"eval_accuracy": 0.383644324092701,
"eval_loss": 3.3621912002563477,
"eval_runtime": 53.322,
"eval_samples_per_second": 337.797,
"eval_steps_per_second": 21.117,
"step": 23000
},
{
"epoch": 6.2029500430663225,
"grad_norm": 0.353242963552475,
"learning_rate": 0.0005073017369058839,
"loss": 3.2304,
"step": 23050
},
{
"epoch": 6.21640826873385,
"grad_norm": 0.3593711256980896,
"learning_rate": 0.0005070997711054262,
"loss": 3.233,
"step": 23100
},
{
"epoch": 6.229866494401378,
"grad_norm": 0.3433372676372528,
"learning_rate": 0.0005068978053049683,
"loss": 3.2408,
"step": 23150
},
{
"epoch": 6.243324720068906,
"grad_norm": 0.3444235026836395,
"learning_rate": 0.0005066958395045105,
"loss": 3.2375,
"step": 23200
},
{
"epoch": 6.256782945736434,
"grad_norm": 0.3466513156890869,
"learning_rate": 0.0005064938737040527,
"loss": 3.2338,
"step": 23250
},
{
"epoch": 6.270241171403962,
"grad_norm": 0.4486019015312195,
"learning_rate": 0.0005062919079035949,
"loss": 3.2367,
"step": 23300
},
{
"epoch": 6.28369939707149,
"grad_norm": 0.35569101572036743,
"learning_rate": 0.0005060899421031372,
"loss": 3.2306,
"step": 23350
},
{
"epoch": 6.297157622739018,
"grad_norm": 0.34310290217399597,
"learning_rate": 0.0005058879763026793,
"loss": 3.2457,
"step": 23400
},
{
"epoch": 6.310615848406546,
"grad_norm": 0.34421324729919434,
"learning_rate": 0.0005056860105022216,
"loss": 3.2488,
"step": 23450
},
{
"epoch": 6.324074074074074,
"grad_norm": 0.34529271721839905,
"learning_rate": 0.0005054840447017637,
"loss": 3.2421,
"step": 23500
},
{
"epoch": 6.337532299741602,
"grad_norm": 0.3405730128288269,
"learning_rate": 0.0005052820789013061,
"loss": 3.2323,
"step": 23550
},
{
"epoch": 6.35099052540913,
"grad_norm": 0.3444826900959015,
"learning_rate": 0.0005050801131008482,
"loss": 3.244,
"step": 23600
},
{
"epoch": 6.364448751076658,
"grad_norm": 0.332755446434021,
"learning_rate": 0.0005048781473003905,
"loss": 3.2356,
"step": 23650
},
{
"epoch": 6.377906976744186,
"grad_norm": 0.3805544376373291,
"learning_rate": 0.0005046761814999326,
"loss": 3.2533,
"step": 23700
},
{
"epoch": 6.3913652024117145,
"grad_norm": 0.35615286231040955,
"learning_rate": 0.0005044742156994749,
"loss": 3.243,
"step": 23750
},
{
"epoch": 6.404823428079242,
"grad_norm": 0.340789258480072,
"learning_rate": 0.0005042722498990171,
"loss": 3.2459,
"step": 23800
},
{
"epoch": 6.41828165374677,
"grad_norm": 0.338225781917572,
"learning_rate": 0.0005040702840985593,
"loss": 3.242,
"step": 23850
},
{
"epoch": 6.431739879414298,
"grad_norm": 0.3471758961677551,
"learning_rate": 0.0005038683182981015,
"loss": 3.2507,
"step": 23900
},
{
"epoch": 6.445198105081826,
"grad_norm": 0.3620913624763489,
"learning_rate": 0.0005036663524976436,
"loss": 3.2475,
"step": 23950
},
{
"epoch": 6.458656330749354,
"grad_norm": 0.3431723415851593,
"learning_rate": 0.0005034643866971859,
"loss": 3.2451,
"step": 24000
},
{
"epoch": 6.458656330749354,
"eval_accuracy": 0.3841879929579454,
"eval_loss": 3.356856346130371,
"eval_runtime": 53.3316,
"eval_samples_per_second": 337.736,
"eval_steps_per_second": 21.113,
"step": 24000
},
{
"epoch": 6.472114556416882,
"grad_norm": 0.36163797974586487,
"learning_rate": 0.0005032624208967281,
"loss": 3.2586,
"step": 24050
},
{
"epoch": 6.48557278208441,
"grad_norm": 0.32522052526474,
"learning_rate": 0.0005030604550962703,
"loss": 3.2529,
"step": 24100
},
{
"epoch": 6.499031007751938,
"grad_norm": 0.3808082044124603,
"learning_rate": 0.0005028584892958125,
"loss": 3.2499,
"step": 24150
},
{
"epoch": 6.5124892334194655,
"grad_norm": 0.33839014172554016,
"learning_rate": 0.0005026565234953547,
"loss": 3.2481,
"step": 24200
},
{
"epoch": 6.525947459086994,
"grad_norm": 0.3400689363479614,
"learning_rate": 0.0005024545576948969,
"loss": 3.2538,
"step": 24250
},
{
"epoch": 6.539405684754522,
"grad_norm": 0.34009531140327454,
"learning_rate": 0.0005022525918944392,
"loss": 3.2564,
"step": 24300
},
{
"epoch": 6.55286391042205,
"grad_norm": 0.3178061842918396,
"learning_rate": 0.0005020506260939813,
"loss": 3.2522,
"step": 24350
},
{
"epoch": 6.566322136089578,
"grad_norm": 0.3306698799133301,
"learning_rate": 0.0005018486602935236,
"loss": 3.2552,
"step": 24400
},
{
"epoch": 6.579780361757106,
"grad_norm": 0.35276076197624207,
"learning_rate": 0.0005016466944930658,
"loss": 3.2585,
"step": 24450
},
{
"epoch": 6.593238587424634,
"grad_norm": 0.34085813164711,
"learning_rate": 0.0005014447286926081,
"loss": 3.2659,
"step": 24500
},
{
"epoch": 6.606696813092162,
"grad_norm": 0.34538936614990234,
"learning_rate": 0.0005012427628921502,
"loss": 3.2513,
"step": 24550
},
{
"epoch": 6.62015503875969,
"grad_norm": 0.33835741877555847,
"learning_rate": 0.0005010407970916925,
"loss": 3.257,
"step": 24600
},
{
"epoch": 6.633613264427218,
"grad_norm": 0.3618541359901428,
"learning_rate": 0.0005008388312912346,
"loss": 3.2474,
"step": 24650
},
{
"epoch": 6.647071490094746,
"grad_norm": 0.3588375449180603,
"learning_rate": 0.0005006368654907769,
"loss": 3.2441,
"step": 24700
},
{
"epoch": 6.660529715762274,
"grad_norm": 0.35771581530570984,
"learning_rate": 0.0005004348996903191,
"loss": 3.2597,
"step": 24750
},
{
"epoch": 6.673987941429802,
"grad_norm": 0.35776013135910034,
"learning_rate": 0.0005002329338898613,
"loss": 3.2605,
"step": 24800
},
{
"epoch": 6.68744616709733,
"grad_norm": 0.35457316040992737,
"learning_rate": 0.0005000309680894035,
"loss": 3.2593,
"step": 24850
},
{
"epoch": 6.7009043927648575,
"grad_norm": 0.3595224916934967,
"learning_rate": 0.0004998290022889456,
"loss": 3.2642,
"step": 24900
},
{
"epoch": 6.714362618432386,
"grad_norm": 0.3422452509403229,
"learning_rate": 0.0004996270364884879,
"loss": 3.2614,
"step": 24950
},
{
"epoch": 6.727820844099914,
"grad_norm": 0.335967093706131,
"learning_rate": 0.0004994250706880301,
"loss": 3.2585,
"step": 25000
},
{
"epoch": 6.727820844099914,
"eval_accuracy": 0.38525447049216593,
"eval_loss": 3.3472113609313965,
"eval_runtime": 53.3326,
"eval_samples_per_second": 337.73,
"eval_steps_per_second": 21.113,
"step": 25000
},
{
"epoch": 6.741279069767442,
"grad_norm": 0.33750924468040466,
"learning_rate": 0.0004992231048875723,
"loss": 3.2587,
"step": 25050
},
{
"epoch": 6.754737295434969,
"grad_norm": 0.3297862708568573,
"learning_rate": 0.0004990211390871145,
"loss": 3.264,
"step": 25100
},
{
"epoch": 6.768195521102498,
"grad_norm": 0.3606378436088562,
"learning_rate": 0.0004988191732866567,
"loss": 3.2574,
"step": 25150
},
{
"epoch": 6.781653746770026,
"grad_norm": 0.3313273787498474,
"learning_rate": 0.0004986172074861989,
"loss": 3.2519,
"step": 25200
},
{
"epoch": 6.795111972437554,
"grad_norm": 0.348251610994339,
"learning_rate": 0.0004984152416857412,
"loss": 3.2619,
"step": 25250
},
{
"epoch": 6.808570198105082,
"grad_norm": 0.33993712067604065,
"learning_rate": 0.0004982132758852834,
"loss": 3.2576,
"step": 25300
},
{
"epoch": 6.822028423772609,
"grad_norm": 0.34746164083480835,
"learning_rate": 0.0004980113100848256,
"loss": 3.2612,
"step": 25350
},
{
"epoch": 6.835486649440138,
"grad_norm": 0.3444181978702545,
"learning_rate": 0.0004978093442843678,
"loss": 3.2617,
"step": 25400
},
{
"epoch": 6.848944875107666,
"grad_norm": 0.3506355583667755,
"learning_rate": 0.0004976073784839101,
"loss": 3.2612,
"step": 25450
},
{
"epoch": 6.862403100775194,
"grad_norm": 0.37607342004776,
"learning_rate": 0.0004974054126834522,
"loss": 3.2593,
"step": 25500
},
{
"epoch": 6.875861326442722,
"grad_norm": 0.33187639713287354,
"learning_rate": 0.0004972034468829945,
"loss": 3.2613,
"step": 25550
},
{
"epoch": 6.8893195521102495,
"grad_norm": 0.33409222960472107,
"learning_rate": 0.0004970014810825366,
"loss": 3.2624,
"step": 25600
},
{
"epoch": 6.902777777777778,
"grad_norm": 0.32660526037216187,
"learning_rate": 0.0004967995152820789,
"loss": 3.2656,
"step": 25650
},
{
"epoch": 6.916236003445306,
"grad_norm": 0.3403673768043518,
"learning_rate": 0.0004965975494816211,
"loss": 3.2674,
"step": 25700
},
{
"epoch": 6.929694229112834,
"grad_norm": 0.3395448327064514,
"learning_rate": 0.0004963955836811633,
"loss": 3.2594,
"step": 25750
},
{
"epoch": 6.943152454780362,
"grad_norm": 0.34910374879837036,
"learning_rate": 0.0004961936178807055,
"loss": 3.2597,
"step": 25800
},
{
"epoch": 6.9566106804478895,
"grad_norm": 0.3608143627643585,
"learning_rate": 0.0004959916520802476,
"loss": 3.2544,
"step": 25850
},
{
"epoch": 6.970068906115418,
"grad_norm": 0.33373093605041504,
"learning_rate": 0.0004957896862797899,
"loss": 3.2658,
"step": 25900
},
{
"epoch": 6.983527131782946,
"grad_norm": 0.36241739988327026,
"learning_rate": 0.0004955877204793321,
"loss": 3.2547,
"step": 25950
},
{
"epoch": 6.996985357450473,
"grad_norm": 0.3471966087818146,
"learning_rate": 0.0004953857546788743,
"loss": 3.2476,
"step": 26000
},
{
"epoch": 6.996985357450473,
"eval_accuracy": 0.3860393353767634,
"eval_loss": 3.3386144638061523,
"eval_runtime": 53.2265,
"eval_samples_per_second": 338.403,
"eval_steps_per_second": 21.155,
"step": 26000
},
{
"epoch": 7.010228251507321,
"grad_norm": 0.32640400528907776,
"learning_rate": 0.0004951837888784165,
"loss": 3.185,
"step": 26050
},
{
"epoch": 7.0236864771748495,
"grad_norm": 0.3508910536766052,
"learning_rate": 0.0004949818230779587,
"loss": 3.1646,
"step": 26100
},
{
"epoch": 7.037144702842378,
"grad_norm": 0.3429078757762909,
"learning_rate": 0.0004947798572775009,
"loss": 3.1603,
"step": 26150
},
{
"epoch": 7.050602928509905,
"grad_norm": 0.31306034326553345,
"learning_rate": 0.0004945778914770432,
"loss": 3.183,
"step": 26200
},
{
"epoch": 7.064061154177433,
"grad_norm": 0.3623344898223877,
"learning_rate": 0.0004943759256765854,
"loss": 3.1838,
"step": 26250
},
{
"epoch": 7.077519379844961,
"grad_norm": 0.3504388630390167,
"learning_rate": 0.0004941739598761276,
"loss": 3.1868,
"step": 26300
},
{
"epoch": 7.09097760551249,
"grad_norm": 0.3301059901714325,
"learning_rate": 0.0004939719940756698,
"loss": 3.1861,
"step": 26350
},
{
"epoch": 7.104435831180017,
"grad_norm": 0.3660713732242584,
"learning_rate": 0.0004937700282752121,
"loss": 3.177,
"step": 26400
},
{
"epoch": 7.117894056847545,
"grad_norm": 0.34936845302581787,
"learning_rate": 0.0004935680624747542,
"loss": 3.1798,
"step": 26450
},
{
"epoch": 7.131352282515073,
"grad_norm": 0.3554314374923706,
"learning_rate": 0.0004933660966742965,
"loss": 3.1774,
"step": 26500
},
{
"epoch": 7.144810508182601,
"grad_norm": 0.3377351760864258,
"learning_rate": 0.0004931641308738386,
"loss": 3.1898,
"step": 26550
},
{
"epoch": 7.15826873385013,
"grad_norm": 0.35610777139663696,
"learning_rate": 0.0004929621650733809,
"loss": 3.1923,
"step": 26600
},
{
"epoch": 7.171726959517657,
"grad_norm": 0.3468087911605835,
"learning_rate": 0.0004927601992729231,
"loss": 3.1869,
"step": 26650
},
{
"epoch": 7.185185185185185,
"grad_norm": 0.34939804673194885,
"learning_rate": 0.0004925582334724652,
"loss": 3.1861,
"step": 26700
},
{
"epoch": 7.198643410852713,
"grad_norm": 0.3704433739185333,
"learning_rate": 0.0004923562676720075,
"loss": 3.1948,
"step": 26750
},
{
"epoch": 7.2121016365202415,
"grad_norm": 0.36642688512802124,
"learning_rate": 0.0004921543018715497,
"loss": 3.1992,
"step": 26800
},
{
"epoch": 7.225559862187769,
"grad_norm": 0.34551697969436646,
"learning_rate": 0.0004919523360710919,
"loss": 3.196,
"step": 26850
},
{
"epoch": 7.239018087855297,
"grad_norm": 0.33420753479003906,
"learning_rate": 0.0004917503702706341,
"loss": 3.2072,
"step": 26900
},
{
"epoch": 7.252476313522825,
"grad_norm": 0.36563876271247864,
"learning_rate": 0.0004915484044701763,
"loss": 3.2002,
"step": 26950
},
{
"epoch": 7.265934539190353,
"grad_norm": 0.3770323693752289,
"learning_rate": 0.0004913464386697185,
"loss": 3.2025,
"step": 27000
},
{
"epoch": 7.265934539190353,
"eval_accuracy": 0.38613451002223786,
"eval_loss": 3.347355604171753,
"eval_runtime": 53.2539,
"eval_samples_per_second": 338.229,
"eval_steps_per_second": 21.144,
"step": 27000
},
{
"epoch": 7.279392764857882,
"grad_norm": 0.34507986903190613,
"learning_rate": 0.0004911444728692608,
"loss": 3.2055,
"step": 27050
},
{
"epoch": 7.292850990525409,
"grad_norm": 0.32827848196029663,
"learning_rate": 0.000490942507068803,
"loss": 3.2089,
"step": 27100
},
{
"epoch": 7.306309216192937,
"grad_norm": 0.38157767057418823,
"learning_rate": 0.0004907405412683452,
"loss": 3.199,
"step": 27150
},
{
"epoch": 7.319767441860465,
"grad_norm": 0.3539191484451294,
"learning_rate": 0.0004905385754678874,
"loss": 3.2145,
"step": 27200
},
{
"epoch": 7.333225667527993,
"grad_norm": 0.37116944789886475,
"learning_rate": 0.0004903366096674297,
"loss": 3.2057,
"step": 27250
},
{
"epoch": 7.346683893195521,
"grad_norm": 0.35318002104759216,
"learning_rate": 0.0004901346438669718,
"loss": 3.203,
"step": 27300
},
{
"epoch": 7.360142118863049,
"grad_norm": 0.3373660147190094,
"learning_rate": 0.0004899326780665141,
"loss": 3.2205,
"step": 27350
},
{
"epoch": 7.373600344530577,
"grad_norm": 0.36426669359207153,
"learning_rate": 0.0004897307122660562,
"loss": 3.2197,
"step": 27400
},
{
"epoch": 7.387058570198105,
"grad_norm": 0.3387640118598938,
"learning_rate": 0.0004895287464655985,
"loss": 3.2181,
"step": 27450
},
{
"epoch": 7.4005167958656335,
"grad_norm": 0.35902735590934753,
"learning_rate": 0.0004893267806651407,
"loss": 3.213,
"step": 27500
},
{
"epoch": 7.413975021533161,
"grad_norm": 0.3508385419845581,
"learning_rate": 0.0004891248148646829,
"loss": 3.205,
"step": 27550
},
{
"epoch": 7.427433247200689,
"grad_norm": 0.36484771966934204,
"learning_rate": 0.0004889228490642251,
"loss": 3.2166,
"step": 27600
},
{
"epoch": 7.440891472868217,
"grad_norm": 0.33635297417640686,
"learning_rate": 0.0004887208832637672,
"loss": 3.2187,
"step": 27650
},
{
"epoch": 7.454349698535745,
"grad_norm": 0.3573185205459595,
"learning_rate": 0.0004885189174633095,
"loss": 3.2118,
"step": 27700
},
{
"epoch": 7.467807924203273,
"grad_norm": 0.33649978041648865,
"learning_rate": 0.0004883169516628517,
"loss": 3.2172,
"step": 27750
},
{
"epoch": 7.481266149870801,
"grad_norm": 0.34777286648750305,
"learning_rate": 0.0004881149858623939,
"loss": 3.2072,
"step": 27800
},
{
"epoch": 7.494724375538329,
"grad_norm": 0.37506943941116333,
"learning_rate": 0.00048791302006193614,
"loss": 3.2176,
"step": 27850
},
{
"epoch": 7.508182601205857,
"grad_norm": 0.32190048694610596,
"learning_rate": 0.00048771105426147833,
"loss": 3.2218,
"step": 27900
},
{
"epoch": 7.521640826873385,
"grad_norm": 0.3466486930847168,
"learning_rate": 0.00048750908846102053,
"loss": 3.2157,
"step": 27950
},
{
"epoch": 7.535099052540913,
"grad_norm": 0.3516653776168823,
"learning_rate": 0.0004873071226605628,
"loss": 3.2282,
"step": 28000
},
{
"epoch": 7.535099052540913,
"eval_accuracy": 0.3867050146608067,
"eval_loss": 3.338799476623535,
"eval_runtime": 53.4413,
"eval_samples_per_second": 337.042,
"eval_steps_per_second": 21.07,
"step": 28000
},
{
"epoch": 7.548557278208441,
"grad_norm": 0.33157870173454285,
"learning_rate": 0.00048710515686010503,
"loss": 3.2251,
"step": 28050
},
{
"epoch": 7.562015503875969,
"grad_norm": 0.3672688901424408,
"learning_rate": 0.0004869031910596472,
"loss": 3.2207,
"step": 28100
},
{
"epoch": 7.575473729543497,
"grad_norm": 0.3594033718109131,
"learning_rate": 0.0004867012252591894,
"loss": 3.2099,
"step": 28150
},
{
"epoch": 7.588931955211025,
"grad_norm": 0.3516235947608948,
"learning_rate": 0.0004864992594587316,
"loss": 3.2308,
"step": 28200
},
{
"epoch": 7.602390180878553,
"grad_norm": 0.33869901299476624,
"learning_rate": 0.0004862972936582738,
"loss": 3.2195,
"step": 28250
},
{
"epoch": 7.615848406546081,
"grad_norm": 0.34088006615638733,
"learning_rate": 0.00048609532785781606,
"loss": 3.2307,
"step": 28300
},
{
"epoch": 7.629306632213609,
"grad_norm": 0.3419373333454132,
"learning_rate": 0.00048589336205735826,
"loss": 3.2092,
"step": 28350
},
{
"epoch": 7.6427648578811365,
"grad_norm": 0.34658679366111755,
"learning_rate": 0.00048569139625690046,
"loss": 3.2286,
"step": 28400
},
{
"epoch": 7.656223083548665,
"grad_norm": 0.3279028534889221,
"learning_rate": 0.00048548943045644265,
"loss": 3.2199,
"step": 28450
},
{
"epoch": 7.669681309216193,
"grad_norm": 0.3251379728317261,
"learning_rate": 0.00048528746465598485,
"loss": 3.2218,
"step": 28500
},
{
"epoch": 7.683139534883721,
"grad_norm": 0.3511503338813782,
"learning_rate": 0.0004850854988555271,
"loss": 3.2342,
"step": 28550
},
{
"epoch": 7.696597760551249,
"grad_norm": 0.3733566105365753,
"learning_rate": 0.0004848835330550693,
"loss": 3.2329,
"step": 28600
},
{
"epoch": 7.7100559862187765,
"grad_norm": 0.33851513266563416,
"learning_rate": 0.0004846815672546115,
"loss": 3.2288,
"step": 28650
},
{
"epoch": 7.723514211886305,
"grad_norm": 0.3473910689353943,
"learning_rate": 0.0004844796014541537,
"loss": 3.2182,
"step": 28700
},
{
"epoch": 7.736972437553833,
"grad_norm": 0.36879199743270874,
"learning_rate": 0.0004842776356536959,
"loss": 3.2237,
"step": 28750
},
{
"epoch": 7.750430663221361,
"grad_norm": 0.3303454518318176,
"learning_rate": 0.00048407566985323813,
"loss": 3.2348,
"step": 28800
},
{
"epoch": 7.763888888888889,
"grad_norm": 0.381592720746994,
"learning_rate": 0.0004838737040527804,
"loss": 3.2268,
"step": 28850
},
{
"epoch": 7.777347114556417,
"grad_norm": 0.36583423614501953,
"learning_rate": 0.0004836717382523226,
"loss": 3.231,
"step": 28900
},
{
"epoch": 7.790805340223945,
"grad_norm": 0.3528934419155121,
"learning_rate": 0.0004834697724518648,
"loss": 3.2303,
"step": 28950
},
{
"epoch": 7.804263565891473,
"grad_norm": 0.3354852795600891,
"learning_rate": 0.000483267806651407,
"loss": 3.2378,
"step": 29000
},
{
"epoch": 7.804263565891473,
"eval_accuracy": 0.38734516193379237,
"eval_loss": 3.33156418800354,
"eval_runtime": 53.2846,
"eval_samples_per_second": 338.034,
"eval_steps_per_second": 21.132,
"step": 29000
},
{
"epoch": 7.817721791559001,
"grad_norm": 0.37304386496543884,
"learning_rate": 0.0004830658408509492,
"loss": 3.2271,
"step": 29050
},
{
"epoch": 7.831180017226529,
"grad_norm": 0.327325701713562,
"learning_rate": 0.0004828638750504914,
"loss": 3.2244,
"step": 29100
},
{
"epoch": 7.844638242894057,
"grad_norm": 0.35642972588539124,
"learning_rate": 0.0004826619092500336,
"loss": 3.2359,
"step": 29150
},
{
"epoch": 7.858096468561585,
"grad_norm": 0.33831554651260376,
"learning_rate": 0.0004824599434495758,
"loss": 3.2359,
"step": 29200
},
{
"epoch": 7.871554694229113,
"grad_norm": 0.33137044310569763,
"learning_rate": 0.00048225797764911806,
"loss": 3.2295,
"step": 29250
},
{
"epoch": 7.885012919896641,
"grad_norm": 0.34678998589515686,
"learning_rate": 0.00048205601184866026,
"loss": 3.2316,
"step": 29300
},
{
"epoch": 7.8984711455641685,
"grad_norm": 0.3184164762496948,
"learning_rate": 0.00048185404604820245,
"loss": 3.24,
"step": 29350
},
{
"epoch": 7.911929371231697,
"grad_norm": 0.3482532799243927,
"learning_rate": 0.00048165208024774465,
"loss": 3.2323,
"step": 29400
},
{
"epoch": 7.925387596899225,
"grad_norm": 0.33416080474853516,
"learning_rate": 0.00048145011444728685,
"loss": 3.2354,
"step": 29450
},
{
"epoch": 7.938845822566753,
"grad_norm": 0.3546711802482605,
"learning_rate": 0.0004812481486468291,
"loss": 3.2299,
"step": 29500
},
{
"epoch": 7.95230404823428,
"grad_norm": 0.3453027307987213,
"learning_rate": 0.0004810461828463713,
"loss": 3.2455,
"step": 29550
},
{
"epoch": 7.965762273901809,
"grad_norm": 0.3534720242023468,
"learning_rate": 0.0004808442170459135,
"loss": 3.2348,
"step": 29600
},
{
"epoch": 7.979220499569337,
"grad_norm": 0.34397897124290466,
"learning_rate": 0.0004806422512454557,
"loss": 3.2336,
"step": 29650
},
{
"epoch": 7.992678725236865,
"grad_norm": 0.35581567883491516,
"learning_rate": 0.00048044028544499793,
"loss": 3.2316,
"step": 29700
},
{
"epoch": 8.005921619293712,
"grad_norm": 0.37843331694602966,
"learning_rate": 0.0004802383196445402,
"loss": 3.1745,
"step": 29750
},
{
"epoch": 8.01937984496124,
"grad_norm": 0.38241565227508545,
"learning_rate": 0.0004800363538440824,
"loss": 3.1428,
"step": 29800
},
{
"epoch": 8.032838070628769,
"grad_norm": 0.382955402135849,
"learning_rate": 0.0004798343880436246,
"loss": 3.1382,
"step": 29850
},
{
"epoch": 8.046296296296296,
"grad_norm": 0.3386884331703186,
"learning_rate": 0.00047963242224316683,
"loss": 3.1443,
"step": 29900
},
{
"epoch": 8.059754521963825,
"grad_norm": 0.3572510778903961,
"learning_rate": 0.000479430456442709,
"loss": 3.1356,
"step": 29950
},
{
"epoch": 8.073212747631352,
"grad_norm": 0.3773634731769562,
"learning_rate": 0.0004792284906422512,
"loss": 3.1434,
"step": 30000
},
{
"epoch": 8.073212747631352,
"eval_accuracy": 0.3869395832219703,
"eval_loss": 3.3406853675842285,
"eval_runtime": 53.3677,
"eval_samples_per_second": 337.508,
"eval_steps_per_second": 21.099,
"step": 30000
},
{
"epoch": 8.08667097329888,
"grad_norm": 0.3504475951194763,
"learning_rate": 0.0004790265248417934,
"loss": 3.148,
"step": 30050
},
{
"epoch": 8.100129198966409,
"grad_norm": 0.3641934096813202,
"learning_rate": 0.0004788245590413356,
"loss": 3.1534,
"step": 30100
},
{
"epoch": 8.113587424633936,
"grad_norm": 0.361492782831192,
"learning_rate": 0.00047862259324087786,
"loss": 3.1513,
"step": 30150
},
{
"epoch": 8.127045650301465,
"grad_norm": 0.3426317572593689,
"learning_rate": 0.00047842062744042006,
"loss": 3.1629,
"step": 30200
},
{
"epoch": 8.140503875968992,
"grad_norm": 0.3521021604537964,
"learning_rate": 0.00047821866163996225,
"loss": 3.164,
"step": 30250
},
{
"epoch": 8.15396210163652,
"grad_norm": 0.34249168634414673,
"learning_rate": 0.00047801669583950445,
"loss": 3.1764,
"step": 30300
},
{
"epoch": 8.167420327304049,
"grad_norm": 0.3548749089241028,
"learning_rate": 0.00047781473003904665,
"loss": 3.1532,
"step": 30350
},
{
"epoch": 8.180878552971576,
"grad_norm": 0.3407946228981018,
"learning_rate": 0.0004776127642385889,
"loss": 3.1637,
"step": 30400
},
{
"epoch": 8.194336778639105,
"grad_norm": 0.347978800535202,
"learning_rate": 0.0004774107984381311,
"loss": 3.174,
"step": 30450
},
{
"epoch": 8.207795004306632,
"grad_norm": 0.36164358258247375,
"learning_rate": 0.0004772088326376733,
"loss": 3.1653,
"step": 30500
},
{
"epoch": 8.22125322997416,
"grad_norm": 0.3417772650718689,
"learning_rate": 0.0004770068668372155,
"loss": 3.1683,
"step": 30550
},
{
"epoch": 8.234711455641689,
"grad_norm": 0.3758831024169922,
"learning_rate": 0.0004768049010367577,
"loss": 3.1801,
"step": 30600
},
{
"epoch": 8.248169681309216,
"grad_norm": 0.3716782033443451,
"learning_rate": 0.0004766029352363,
"loss": 3.1733,
"step": 30650
},
{
"epoch": 8.261627906976745,
"grad_norm": 0.3527848422527313,
"learning_rate": 0.0004764009694358422,
"loss": 3.1789,
"step": 30700
},
{
"epoch": 8.275086132644272,
"grad_norm": 0.3328978419303894,
"learning_rate": 0.0004761990036353844,
"loss": 3.1694,
"step": 30750
},
{
"epoch": 8.2885443583118,
"grad_norm": 0.3273838758468628,
"learning_rate": 0.0004759970378349266,
"loss": 3.1767,
"step": 30800
},
{
"epoch": 8.302002583979329,
"grad_norm": 0.35219597816467285,
"learning_rate": 0.0004757950720344688,
"loss": 3.1719,
"step": 30850
},
{
"epoch": 8.315460809646856,
"grad_norm": 0.33684444427490234,
"learning_rate": 0.000475593106234011,
"loss": 3.1786,
"step": 30900
},
{
"epoch": 8.328919035314383,
"grad_norm": 0.3632763922214508,
"learning_rate": 0.0004753911404335532,
"loss": 3.185,
"step": 30950
},
{
"epoch": 8.342377260981912,
"grad_norm": 0.3545039892196655,
"learning_rate": 0.0004751891746330954,
"loss": 3.1867,
"step": 31000
},
{
"epoch": 8.342377260981912,
"eval_accuracy": 0.3874767332758809,
"eval_loss": 3.337397575378418,
"eval_runtime": 53.4419,
"eval_samples_per_second": 337.039,
"eval_steps_per_second": 21.07,
"step": 31000
},
{
"epoch": 8.35583548664944,
"grad_norm": 0.3302076756954193,
"learning_rate": 0.0004749872088326376,
"loss": 3.1899,
"step": 31050
},
{
"epoch": 8.369293712316969,
"grad_norm": 0.33575427532196045,
"learning_rate": 0.00047478524303217986,
"loss": 3.1839,
"step": 31100
},
{
"epoch": 8.382751937984496,
"grad_norm": 0.35871943831443787,
"learning_rate": 0.00047458327723172206,
"loss": 3.1823,
"step": 31150
},
{
"epoch": 8.396210163652023,
"grad_norm": 0.357995867729187,
"learning_rate": 0.00047438131143126425,
"loss": 3.1952,
"step": 31200
},
{
"epoch": 8.409668389319553,
"grad_norm": 0.37615224719047546,
"learning_rate": 0.00047417934563080645,
"loss": 3.1825,
"step": 31250
},
{
"epoch": 8.42312661498708,
"grad_norm": 0.34395626187324524,
"learning_rate": 0.00047397737983034864,
"loss": 3.1975,
"step": 31300
},
{
"epoch": 8.436584840654609,
"grad_norm": 0.3637622892856598,
"learning_rate": 0.0004737754140298909,
"loss": 3.1957,
"step": 31350
},
{
"epoch": 8.450043066322136,
"grad_norm": 0.33689847588539124,
"learning_rate": 0.0004735734482294331,
"loss": 3.1863,
"step": 31400
},
{
"epoch": 8.463501291989663,
"grad_norm": 0.3541131913661957,
"learning_rate": 0.0004733714824289753,
"loss": 3.1859,
"step": 31450
},
{
"epoch": 8.476959517657193,
"grad_norm": 0.33903294801712036,
"learning_rate": 0.0004731695166285175,
"loss": 3.1894,
"step": 31500
},
{
"epoch": 8.49041774332472,
"grad_norm": 0.3368317782878876,
"learning_rate": 0.0004729675508280598,
"loss": 3.1925,
"step": 31550
},
{
"epoch": 8.503875968992247,
"grad_norm": 0.34673404693603516,
"learning_rate": 0.000472765585027602,
"loss": 3.1997,
"step": 31600
},
{
"epoch": 8.517334194659776,
"grad_norm": 0.3494406044483185,
"learning_rate": 0.0004725636192271442,
"loss": 3.1924,
"step": 31650
},
{
"epoch": 8.530792420327304,
"grad_norm": 0.34491440653800964,
"learning_rate": 0.0004723616534266864,
"loss": 3.1889,
"step": 31700
},
{
"epoch": 8.544250645994833,
"grad_norm": 0.3356218934059143,
"learning_rate": 0.0004721596876262286,
"loss": 3.1988,
"step": 31750
},
{
"epoch": 8.55770887166236,
"grad_norm": 0.3772531747817993,
"learning_rate": 0.0004719577218257708,
"loss": 3.199,
"step": 31800
},
{
"epoch": 8.571167097329887,
"grad_norm": 0.36063918471336365,
"learning_rate": 0.000471755756025313,
"loss": 3.2036,
"step": 31850
},
{
"epoch": 8.584625322997416,
"grad_norm": 0.37636640667915344,
"learning_rate": 0.0004715537902248552,
"loss": 3.1986,
"step": 31900
},
{
"epoch": 8.598083548664944,
"grad_norm": 0.35377389192581177,
"learning_rate": 0.0004713518244243974,
"loss": 3.1894,
"step": 31950
},
{
"epoch": 8.611541774332473,
"grad_norm": 0.35772061347961426,
"learning_rate": 0.00047114985862393966,
"loss": 3.1894,
"step": 32000
},
{
"epoch": 8.611541774332473,
"eval_accuracy": 0.38834927617291887,
"eval_loss": 3.3256490230560303,
"eval_runtime": 53.5289,
"eval_samples_per_second": 336.491,
"eval_steps_per_second": 21.035,
"step": 32000
},
{
"epoch": 8.625,
"grad_norm": 0.3695344924926758,
"learning_rate": 0.00047094789282348186,
"loss": 3.1975,
"step": 32050
},
{
"epoch": 8.638458225667527,
"grad_norm": 0.4119062125682831,
"learning_rate": 0.00047074592702302405,
"loss": 3.1931,
"step": 32100
},
{
"epoch": 8.651916451335056,
"grad_norm": 0.35740119218826294,
"learning_rate": 0.00047054396122256625,
"loss": 3.1945,
"step": 32150
},
{
"epoch": 8.665374677002584,
"grad_norm": 0.3605038821697235,
"learning_rate": 0.00047034199542210844,
"loss": 3.1909,
"step": 32200
},
{
"epoch": 8.678832902670113,
"grad_norm": 0.3636098802089691,
"learning_rate": 0.0004701400296216507,
"loss": 3.2043,
"step": 32250
},
{
"epoch": 8.69229112833764,
"grad_norm": 0.3707864582538605,
"learning_rate": 0.0004699380638211929,
"loss": 3.2117,
"step": 32300
},
{
"epoch": 8.705749354005167,
"grad_norm": 0.34066513180732727,
"learning_rate": 0.0004697360980207351,
"loss": 3.2029,
"step": 32350
},
{
"epoch": 8.719207579672696,
"grad_norm": 0.34851470589637756,
"learning_rate": 0.0004695341322202773,
"loss": 3.1863,
"step": 32400
},
{
"epoch": 8.732665805340224,
"grad_norm": 0.35019055008888245,
"learning_rate": 0.0004693321664198196,
"loss": 3.2009,
"step": 32450
},
{
"epoch": 8.746124031007753,
"grad_norm": 0.35675832629203796,
"learning_rate": 0.0004691302006193618,
"loss": 3.1981,
"step": 32500
},
{
"epoch": 8.75958225667528,
"grad_norm": 0.34505200386047363,
"learning_rate": 0.000468928234818904,
"loss": 3.2123,
"step": 32550
},
{
"epoch": 8.773040482342807,
"grad_norm": 0.3546963632106781,
"learning_rate": 0.0004687262690184462,
"loss": 3.2002,
"step": 32600
},
{
"epoch": 8.786498708010337,
"grad_norm": 0.3934785723686218,
"learning_rate": 0.00046852430321798837,
"loss": 3.2091,
"step": 32650
},
{
"epoch": 8.799956933677864,
"grad_norm": 0.3397079408168793,
"learning_rate": 0.0004683223374175306,
"loss": 3.197,
"step": 32700
},
{
"epoch": 8.813415159345391,
"grad_norm": 0.3354127109050751,
"learning_rate": 0.0004681203716170728,
"loss": 3.2114,
"step": 32750
},
{
"epoch": 8.82687338501292,
"grad_norm": 0.35856249928474426,
"learning_rate": 0.000467918405816615,
"loss": 3.2018,
"step": 32800
},
{
"epoch": 8.840331610680447,
"grad_norm": 0.34060561656951904,
"learning_rate": 0.0004677164400161572,
"loss": 3.2029,
"step": 32850
},
{
"epoch": 8.853789836347977,
"grad_norm": 0.3674628734588623,
"learning_rate": 0.0004675144742156994,
"loss": 3.2096,
"step": 32900
},
{
"epoch": 8.867248062015504,
"grad_norm": 0.3439692258834839,
"learning_rate": 0.00046731250841524166,
"loss": 3.2028,
"step": 32950
},
{
"epoch": 8.880706287683031,
"grad_norm": 0.3598936200141907,
"learning_rate": 0.00046711054261478385,
"loss": 3.2085,
"step": 33000
},
{
"epoch": 8.880706287683031,
"eval_accuracy": 0.38856559206234764,
"eval_loss": 3.3211824893951416,
"eval_runtime": 53.2603,
"eval_samples_per_second": 338.188,
"eval_steps_per_second": 21.141,
"step": 33000
},
{
"epoch": 8.89416451335056,
"grad_norm": 0.3437979817390442,
"learning_rate": 0.00046690857681432605,
"loss": 3.1999,
"step": 33050
},
{
"epoch": 8.907622739018088,
"grad_norm": 0.3484727740287781,
"learning_rate": 0.00046670661101386825,
"loss": 3.1986,
"step": 33100
},
{
"epoch": 8.921080964685617,
"grad_norm": 0.35170236229896545,
"learning_rate": 0.00046650464521341044,
"loss": 3.2048,
"step": 33150
},
{
"epoch": 8.934539190353144,
"grad_norm": 0.3486970067024231,
"learning_rate": 0.0004663026794129527,
"loss": 3.2057,
"step": 33200
},
{
"epoch": 8.947997416020671,
"grad_norm": 0.33753907680511475,
"learning_rate": 0.0004661007136124949,
"loss": 3.21,
"step": 33250
},
{
"epoch": 8.9614556416882,
"grad_norm": 0.3692246675491333,
"learning_rate": 0.00046589874781203714,
"loss": 3.2023,
"step": 33300
},
{
"epoch": 8.974913867355728,
"grad_norm": 0.3209538757801056,
"learning_rate": 0.0004656967820115794,
"loss": 3.1987,
"step": 33350
},
{
"epoch": 8.988372093023255,
"grad_norm": 0.39224815368652344,
"learning_rate": 0.0004654948162111216,
"loss": 3.2082,
"step": 33400
},
{
"epoch": 9.001614987080103,
"grad_norm": 0.35652458667755127,
"learning_rate": 0.0004652928504106638,
"loss": 3.2015,
"step": 33450
},
{
"epoch": 9.015073212747632,
"grad_norm": 0.3708120584487915,
"learning_rate": 0.000465090884610206,
"loss": 3.1127,
"step": 33500
},
{
"epoch": 9.02853143841516,
"grad_norm": 0.34585344791412354,
"learning_rate": 0.00046488891880974817,
"loss": 3.1111,
"step": 33550
},
{
"epoch": 9.041989664082687,
"grad_norm": 0.36985328793525696,
"learning_rate": 0.0004646869530092904,
"loss": 3.1147,
"step": 33600
},
{
"epoch": 9.055447889750216,
"grad_norm": 0.3506423532962799,
"learning_rate": 0.0004644849872088326,
"loss": 3.1262,
"step": 33650
},
{
"epoch": 9.068906115417743,
"grad_norm": 0.34678879380226135,
"learning_rate": 0.0004642830214083748,
"loss": 3.1178,
"step": 33700
},
{
"epoch": 9.082364341085272,
"grad_norm": 0.3602910041809082,
"learning_rate": 0.000464081055607917,
"loss": 3.126,
"step": 33750
},
{
"epoch": 9.0958225667528,
"grad_norm": 0.3410084545612335,
"learning_rate": 0.0004638790898074592,
"loss": 3.1358,
"step": 33800
},
{
"epoch": 9.109280792420327,
"grad_norm": 0.36710256338119507,
"learning_rate": 0.00046367712400700146,
"loss": 3.1225,
"step": 33850
},
{
"epoch": 9.122739018087856,
"grad_norm": 0.3548193573951721,
"learning_rate": 0.00046347515820654365,
"loss": 3.1226,
"step": 33900
},
{
"epoch": 9.136197243755383,
"grad_norm": 0.3377094268798828,
"learning_rate": 0.00046327319240608585,
"loss": 3.1227,
"step": 33950
},
{
"epoch": 9.14965546942291,
"grad_norm": 0.34470683336257935,
"learning_rate": 0.00046307122660562805,
"loss": 3.1266,
"step": 34000
},
{
"epoch": 9.14965546942291,
"eval_accuracy": 0.3885327807119672,
"eval_loss": 3.327869415283203,
"eval_runtime": 53.7749,
"eval_samples_per_second": 334.952,
"eval_steps_per_second": 20.939,
"step": 34000
},
{
"epoch": 9.16311369509044,
"grad_norm": 0.360047847032547,
"learning_rate": 0.00046286926080517024,
"loss": 3.1376,
"step": 34050
},
{
"epoch": 9.176571920757967,
"grad_norm": 0.3523208200931549,
"learning_rate": 0.0004626672950047125,
"loss": 3.1269,
"step": 34100
},
{
"epoch": 9.190030146425496,
"grad_norm": 0.3531096279621124,
"learning_rate": 0.0004624653292042547,
"loss": 3.1298,
"step": 34150
},
{
"epoch": 9.203488372093023,
"grad_norm": 0.33584877848625183,
"learning_rate": 0.00046226336340379694,
"loss": 3.1504,
"step": 34200
},
{
"epoch": 9.21694659776055,
"grad_norm": 0.3406696617603302,
"learning_rate": 0.00046206139760333913,
"loss": 3.1462,
"step": 34250
},
{
"epoch": 9.23040482342808,
"grad_norm": 0.3635725975036621,
"learning_rate": 0.0004618594318028814,
"loss": 3.1605,
"step": 34300
},
{
"epoch": 9.243863049095607,
"grad_norm": 0.35520845651626587,
"learning_rate": 0.0004616574660024236,
"loss": 3.1466,
"step": 34350
},
{
"epoch": 9.257321274763136,
"grad_norm": 0.36953485012054443,
"learning_rate": 0.0004614555002019658,
"loss": 3.1521,
"step": 34400
},
{
"epoch": 9.270779500430663,
"grad_norm": 0.3408142328262329,
"learning_rate": 0.000461253534401508,
"loss": 3.1451,
"step": 34450
},
{
"epoch": 9.28423772609819,
"grad_norm": 0.35455116629600525,
"learning_rate": 0.00046105156860105017,
"loss": 3.1512,
"step": 34500
},
{
"epoch": 9.29769595176572,
"grad_norm": 0.3463025689125061,
"learning_rate": 0.0004608496028005924,
"loss": 3.1463,
"step": 34550
},
{
"epoch": 9.311154177433247,
"grad_norm": 0.35179778933525085,
"learning_rate": 0.0004606476370001346,
"loss": 3.1647,
"step": 34600
},
{
"epoch": 9.324612403100776,
"grad_norm": 0.34766823053359985,
"learning_rate": 0.0004604456711996768,
"loss": 3.154,
"step": 34650
},
{
"epoch": 9.338070628768303,
"grad_norm": 0.3571286201477051,
"learning_rate": 0.000460243705399219,
"loss": 3.1697,
"step": 34700
},
{
"epoch": 9.35152885443583,
"grad_norm": 0.3302684426307678,
"learning_rate": 0.0004600417395987612,
"loss": 3.155,
"step": 34750
},
{
"epoch": 9.36498708010336,
"grad_norm": 0.3882899582386017,
"learning_rate": 0.00045983977379830345,
"loss": 3.155,
"step": 34800
},
{
"epoch": 9.378445305770887,
"grad_norm": 0.3561243414878845,
"learning_rate": 0.00045963780799784565,
"loss": 3.1591,
"step": 34850
},
{
"epoch": 9.391903531438416,
"grad_norm": 0.3711599111557007,
"learning_rate": 0.00045943584219738785,
"loss": 3.1584,
"step": 34900
},
{
"epoch": 9.405361757105943,
"grad_norm": 0.4065340459346771,
"learning_rate": 0.00045923387639693004,
"loss": 3.1684,
"step": 34950
},
{
"epoch": 9.41881998277347,
"grad_norm": 0.3408617675304413,
"learning_rate": 0.00045903191059647224,
"loss": 3.1647,
"step": 35000
},
{
"epoch": 9.41881998277347,
"eval_accuracy": 0.388881537118329,
"eval_loss": 3.3228235244750977,
"eval_runtime": 53.5953,
"eval_samples_per_second": 336.074,
"eval_steps_per_second": 21.009,
"step": 35000
},
{
"epoch": 9.432278208441,
"grad_norm": 0.39076054096221924,
"learning_rate": 0.0004588299447960145,
"loss": 3.1743,
"step": 35050
},
{
"epoch": 9.445736434108527,
"grad_norm": 0.3773688077926636,
"learning_rate": 0.00045862797899555674,
"loss": 3.1596,
"step": 35100
},
{
"epoch": 9.459194659776054,
"grad_norm": 0.33396244049072266,
"learning_rate": 0.00045842601319509894,
"loss": 3.1774,
"step": 35150
},
{
"epoch": 9.472652885443583,
"grad_norm": 0.33974024653434753,
"learning_rate": 0.00045822404739464113,
"loss": 3.1719,
"step": 35200
},
{
"epoch": 9.48611111111111,
"grad_norm": 0.3999760150909424,
"learning_rate": 0.0004580220815941834,
"loss": 3.1653,
"step": 35250
},
{
"epoch": 9.49956933677864,
"grad_norm": 0.3510672450065613,
"learning_rate": 0.0004578201157937256,
"loss": 3.1631,
"step": 35300
},
{
"epoch": 9.513027562446167,
"grad_norm": 0.343685507774353,
"learning_rate": 0.0004576181499932678,
"loss": 3.1687,
"step": 35350
},
{
"epoch": 9.526485788113694,
"grad_norm": 0.34639087319374084,
"learning_rate": 0.00045741618419280997,
"loss": 3.1564,
"step": 35400
},
{
"epoch": 9.539944013781223,
"grad_norm": 0.37675940990448,
"learning_rate": 0.0004572142183923522,
"loss": 3.1699,
"step": 35450
},
{
"epoch": 9.55340223944875,
"grad_norm": 0.35573285818099976,
"learning_rate": 0.0004570122525918944,
"loss": 3.1742,
"step": 35500
},
{
"epoch": 9.56686046511628,
"grad_norm": 0.3698192834854126,
"learning_rate": 0.0004568102867914366,
"loss": 3.1801,
"step": 35550
},
{
"epoch": 9.580318690783807,
"grad_norm": 0.3495495319366455,
"learning_rate": 0.0004566083209909788,
"loss": 3.1716,
"step": 35600
},
{
"epoch": 9.593776916451334,
"grad_norm": 0.3501206040382385,
"learning_rate": 0.000456406355190521,
"loss": 3.1683,
"step": 35650
},
{
"epoch": 9.607235142118864,
"grad_norm": 0.38374894857406616,
"learning_rate": 0.00045620438939006326,
"loss": 3.179,
"step": 35700
},
{
"epoch": 9.62069336778639,
"grad_norm": 0.34734368324279785,
"learning_rate": 0.00045600242358960545,
"loss": 3.1827,
"step": 35750
},
{
"epoch": 9.634151593453918,
"grad_norm": 0.33780547976493835,
"learning_rate": 0.00045580045778914765,
"loss": 3.1679,
"step": 35800
},
{
"epoch": 9.647609819121447,
"grad_norm": 0.35860759019851685,
"learning_rate": 0.00045559849198868984,
"loss": 3.1798,
"step": 35850
},
{
"epoch": 9.661068044788975,
"grad_norm": 0.3601613938808441,
"learning_rate": 0.00045539652618823204,
"loss": 3.1827,
"step": 35900
},
{
"epoch": 9.674526270456504,
"grad_norm": 0.35015615820884705,
"learning_rate": 0.0004551945603877743,
"loss": 3.1855,
"step": 35950
},
{
"epoch": 9.687984496124031,
"grad_norm": 0.3426166772842407,
"learning_rate": 0.00045499259458731654,
"loss": 3.1814,
"step": 36000
},
{
"epoch": 9.687984496124031,
"eval_accuracy": 0.38963543764909064,
"eval_loss": 3.3155407905578613,
"eval_runtime": 53.4256,
"eval_samples_per_second": 337.142,
"eval_steps_per_second": 21.076,
"step": 36000
},
{
"epoch": 9.701442721791558,
"grad_norm": 0.3451278805732727,
"learning_rate": 0.00045479062878685874,
"loss": 3.188,
"step": 36050
},
{
"epoch": 9.714900947459087,
"grad_norm": 0.368107408285141,
"learning_rate": 0.00045458866298640093,
"loss": 3.1815,
"step": 36100
},
{
"epoch": 9.728359173126615,
"grad_norm": 0.3657575249671936,
"learning_rate": 0.0004543866971859432,
"loss": 3.1868,
"step": 36150
},
{
"epoch": 9.741817398794144,
"grad_norm": 0.34643518924713135,
"learning_rate": 0.0004541847313854854,
"loss": 3.1751,
"step": 36200
},
{
"epoch": 9.755275624461671,
"grad_norm": 0.36216220259666443,
"learning_rate": 0.0004539827655850276,
"loss": 3.1826,
"step": 36250
},
{
"epoch": 9.768733850129198,
"grad_norm": 0.3650382161140442,
"learning_rate": 0.00045378079978456977,
"loss": 3.1718,
"step": 36300
},
{
"epoch": 9.782192075796727,
"grad_norm": 0.348186194896698,
"learning_rate": 0.00045357883398411197,
"loss": 3.1654,
"step": 36350
},
{
"epoch": 9.795650301464255,
"grad_norm": 0.3629589378833771,
"learning_rate": 0.0004533768681836542,
"loss": 3.1869,
"step": 36400
},
{
"epoch": 9.809108527131784,
"grad_norm": 0.3417031764984131,
"learning_rate": 0.0004531749023831964,
"loss": 3.1829,
"step": 36450
},
{
"epoch": 9.822566752799311,
"grad_norm": 0.34222733974456787,
"learning_rate": 0.0004529729365827386,
"loss": 3.191,
"step": 36500
},
{
"epoch": 9.836024978466838,
"grad_norm": 0.341671884059906,
"learning_rate": 0.0004527709707822808,
"loss": 3.189,
"step": 36550
},
{
"epoch": 9.849483204134367,
"grad_norm": 0.3711172938346863,
"learning_rate": 0.000452569004981823,
"loss": 3.1764,
"step": 36600
},
{
"epoch": 9.862941429801895,
"grad_norm": 0.3440135717391968,
"learning_rate": 0.00045236703918136525,
"loss": 3.1641,
"step": 36650
},
{
"epoch": 9.876399655469424,
"grad_norm": 0.3357332646846771,
"learning_rate": 0.00045216507338090745,
"loss": 3.177,
"step": 36700
},
{
"epoch": 9.889857881136951,
"grad_norm": 0.3803945779800415,
"learning_rate": 0.00045196310758044964,
"loss": 3.1844,
"step": 36750
},
{
"epoch": 9.903316106804478,
"grad_norm": 0.3620223104953766,
"learning_rate": 0.00045176114177999184,
"loss": 3.1773,
"step": 36800
},
{
"epoch": 9.916774332472007,
"grad_norm": 0.3589492738246918,
"learning_rate": 0.00045155917597953404,
"loss": 3.1858,
"step": 36850
},
{
"epoch": 9.930232558139535,
"grad_norm": 0.3606046140193939,
"learning_rate": 0.00045135721017907634,
"loss": 3.1901,
"step": 36900
},
{
"epoch": 9.943690783807062,
"grad_norm": 0.3455662429332733,
"learning_rate": 0.00045115524437861854,
"loss": 3.1913,
"step": 36950
},
{
"epoch": 9.957149009474591,
"grad_norm": 0.33837249875068665,
"learning_rate": 0.00045095327857816073,
"loss": 3.1885,
"step": 37000
},
{
"epoch": 9.957149009474591,
"eval_accuracy": 0.3903870565958854,
"eval_loss": 3.3125293254852295,
"eval_runtime": 53.6286,
"eval_samples_per_second": 335.866,
"eval_steps_per_second": 20.996,
"step": 37000
},
{
"epoch": 9.970607235142118,
"grad_norm": 0.37519270181655884,
"learning_rate": 0.00045075131277770293,
"loss": 3.1824,
"step": 37050
},
{
"epoch": 9.984065460809648,
"grad_norm": 0.3906201124191284,
"learning_rate": 0.0004505493469772452,
"loss": 3.1738,
"step": 37100
},
{
"epoch": 9.997523686477175,
"grad_norm": 0.36060965061187744,
"learning_rate": 0.0004503473811767874,
"loss": 3.1862,
"step": 37150
},
{
"epoch": 10.010766580534023,
"grad_norm": 0.35172829031944275,
"learning_rate": 0.00045014541537632957,
"loss": 3.1091,
"step": 37200
},
{
"epoch": 10.02422480620155,
"grad_norm": 0.38957053422927856,
"learning_rate": 0.00044994344957587177,
"loss": 3.0934,
"step": 37250
},
{
"epoch": 10.037683031869078,
"grad_norm": 0.3552197515964508,
"learning_rate": 0.00044974148377541396,
"loss": 3.0892,
"step": 37300
},
{
"epoch": 10.051141257536607,
"grad_norm": 0.39891138672828674,
"learning_rate": 0.0004495395179749562,
"loss": 3.0886,
"step": 37350
},
{
"epoch": 10.064599483204134,
"grad_norm": 0.4051150977611542,
"learning_rate": 0.0004493375521744984,
"loss": 3.1042,
"step": 37400
},
{
"epoch": 10.078057708871663,
"grad_norm": 0.3359229266643524,
"learning_rate": 0.0004491355863740406,
"loss": 3.1038,
"step": 37450
},
{
"epoch": 10.09151593453919,
"grad_norm": 0.35439833998680115,
"learning_rate": 0.0004489336205735828,
"loss": 3.1125,
"step": 37500
},
{
"epoch": 10.104974160206718,
"grad_norm": 0.34396088123321533,
"learning_rate": 0.00044873165477312505,
"loss": 3.108,
"step": 37550
},
{
"epoch": 10.118432385874247,
"grad_norm": 0.3561412990093231,
"learning_rate": 0.00044852968897266725,
"loss": 3.1052,
"step": 37600
},
{
"epoch": 10.131890611541774,
"grad_norm": 0.3590238690376282,
"learning_rate": 0.00044832772317220945,
"loss": 3.1127,
"step": 37650
},
{
"epoch": 10.145348837209303,
"grad_norm": 0.3750346302986145,
"learning_rate": 0.00044812575737175164,
"loss": 3.1132,
"step": 37700
},
{
"epoch": 10.15880706287683,
"grad_norm": 0.3595349192619324,
"learning_rate": 0.00044792379157129384,
"loss": 3.1089,
"step": 37750
},
{
"epoch": 10.172265288544358,
"grad_norm": 0.3501310646533966,
"learning_rate": 0.00044772182577083614,
"loss": 3.1126,
"step": 37800
},
{
"epoch": 10.185723514211887,
"grad_norm": 0.3545890152454376,
"learning_rate": 0.00044751985997037834,
"loss": 3.1221,
"step": 37850
},
{
"epoch": 10.199181739879414,
"grad_norm": 0.3811264932155609,
"learning_rate": 0.00044731789416992053,
"loss": 3.1247,
"step": 37900
},
{
"epoch": 10.212639965546943,
"grad_norm": 0.3635217845439911,
"learning_rate": 0.00044711592836946273,
"loss": 3.1175,
"step": 37950
},
{
"epoch": 10.22609819121447,
"grad_norm": 0.36193686723709106,
"learning_rate": 0.000446913962569005,
"loss": 3.1233,
"step": 38000
},
{
"epoch": 10.22609819121447,
"eval_accuracy": 0.3899182454141249,
"eval_loss": 3.3228235244750977,
"eval_runtime": 53.3214,
"eval_samples_per_second": 337.8,
"eval_steps_per_second": 21.117,
"step": 38000
},
{
"epoch": 10.239556416881998,
"grad_norm": 0.3329092264175415,
"learning_rate": 0.0004467119967685472,
"loss": 3.1267,
"step": 38050
},
{
"epoch": 10.253014642549527,
"grad_norm": 0.3588671386241913,
"learning_rate": 0.00044651003096808937,
"loss": 3.1297,
"step": 38100
},
{
"epoch": 10.266472868217054,
"grad_norm": 0.36286869645118713,
"learning_rate": 0.00044630806516763157,
"loss": 3.1288,
"step": 38150
},
{
"epoch": 10.279931093884581,
"grad_norm": 0.36928683519363403,
"learning_rate": 0.00044610609936717376,
"loss": 3.1187,
"step": 38200
},
{
"epoch": 10.29338931955211,
"grad_norm": 0.34665945172309875,
"learning_rate": 0.000445904133566716,
"loss": 3.1299,
"step": 38250
},
{
"epoch": 10.306847545219638,
"grad_norm": 0.4086586534976959,
"learning_rate": 0.0004457021677662582,
"loss": 3.1397,
"step": 38300
},
{
"epoch": 10.320305770887167,
"grad_norm": 0.3424471318721771,
"learning_rate": 0.0004455002019658004,
"loss": 3.1356,
"step": 38350
},
{
"epoch": 10.333763996554694,
"grad_norm": 0.37157419323921204,
"learning_rate": 0.0004452982361653426,
"loss": 3.1401,
"step": 38400
},
{
"epoch": 10.347222222222221,
"grad_norm": 0.37454769015312195,
"learning_rate": 0.0004450962703648848,
"loss": 3.1251,
"step": 38450
},
{
"epoch": 10.36068044788975,
"grad_norm": 0.3706095218658447,
"learning_rate": 0.00044489430456442705,
"loss": 3.1319,
"step": 38500
},
{
"epoch": 10.374138673557278,
"grad_norm": 0.33369627594947815,
"learning_rate": 0.00044469233876396925,
"loss": 3.1416,
"step": 38550
},
{
"epoch": 10.387596899224807,
"grad_norm": 0.3583381175994873,
"learning_rate": 0.00044449037296351144,
"loss": 3.1457,
"step": 38600
},
{
"epoch": 10.401055124892334,
"grad_norm": 0.39620569348335266,
"learning_rate": 0.0004442884071630537,
"loss": 3.147,
"step": 38650
},
{
"epoch": 10.414513350559861,
"grad_norm": 0.38012731075286865,
"learning_rate": 0.00044408644136259594,
"loss": 3.1386,
"step": 38700
},
{
"epoch": 10.42797157622739,
"grad_norm": 0.3489794135093689,
"learning_rate": 0.00044388447556213814,
"loss": 3.1426,
"step": 38750
},
{
"epoch": 10.441429801894918,
"grad_norm": 0.37140825390815735,
"learning_rate": 0.00044368250976168033,
"loss": 3.14,
"step": 38800
},
{
"epoch": 10.454888027562447,
"grad_norm": 0.3544456362724304,
"learning_rate": 0.00044348054396122253,
"loss": 3.1421,
"step": 38850
},
{
"epoch": 10.468346253229974,
"grad_norm": 0.3525961935520172,
"learning_rate": 0.0004432785781607647,
"loss": 3.1414,
"step": 38900
},
{
"epoch": 10.481804478897502,
"grad_norm": 0.3718273937702179,
"learning_rate": 0.000443076612360307,
"loss": 3.1555,
"step": 38950
},
{
"epoch": 10.49526270456503,
"grad_norm": 0.357573926448822,
"learning_rate": 0.0004428746465598492,
"loss": 3.1452,
"step": 39000
},
{
"epoch": 10.49526270456503,
"eval_accuracy": 0.39027265145697604,
"eval_loss": 3.31387996673584,
"eval_runtime": 53.4319,
"eval_samples_per_second": 337.102,
"eval_steps_per_second": 21.074,
"step": 39000
},
{
"epoch": 10.508720930232558,
"grad_norm": 0.3596917688846588,
"learning_rate": 0.00044267268075939137,
"loss": 3.1411,
"step": 39050
},
{
"epoch": 10.522179155900087,
"grad_norm": 0.3924131989479065,
"learning_rate": 0.00044247071495893357,
"loss": 3.1423,
"step": 39100
},
{
"epoch": 10.535637381567614,
"grad_norm": 0.3612106442451477,
"learning_rate": 0.00044226874915847576,
"loss": 3.1432,
"step": 39150
},
{
"epoch": 10.549095607235142,
"grad_norm": 0.33442962169647217,
"learning_rate": 0.000442066783358018,
"loss": 3.1483,
"step": 39200
},
{
"epoch": 10.56255383290267,
"grad_norm": 0.3581370413303375,
"learning_rate": 0.0004418648175575602,
"loss": 3.1499,
"step": 39250
},
{
"epoch": 10.576012058570198,
"grad_norm": 0.3770648241043091,
"learning_rate": 0.0004416628517571024,
"loss": 3.1449,
"step": 39300
},
{
"epoch": 10.589470284237725,
"grad_norm": 0.3834937512874603,
"learning_rate": 0.0004414608859566446,
"loss": 3.1493,
"step": 39350
},
{
"epoch": 10.602928509905254,
"grad_norm": 0.3559904098510742,
"learning_rate": 0.00044125892015618685,
"loss": 3.1563,
"step": 39400
},
{
"epoch": 10.616386735572782,
"grad_norm": 0.3795510232448578,
"learning_rate": 0.00044105695435572905,
"loss": 3.1429,
"step": 39450
},
{
"epoch": 10.62984496124031,
"grad_norm": 0.3509112000465393,
"learning_rate": 0.00044085498855527124,
"loss": 3.155,
"step": 39500
},
{
"epoch": 10.643303186907838,
"grad_norm": 0.382457971572876,
"learning_rate": 0.0004406530227548135,
"loss": 3.1448,
"step": 39550
},
{
"epoch": 10.656761412575365,
"grad_norm": 0.3792799413204193,
"learning_rate": 0.00044045105695435574,
"loss": 3.1601,
"step": 39600
},
{
"epoch": 10.670219638242894,
"grad_norm": 0.3273200988769531,
"learning_rate": 0.00044024909115389794,
"loss": 3.1474,
"step": 39650
},
{
"epoch": 10.683677863910422,
"grad_norm": 0.3698997497558594,
"learning_rate": 0.00044004712535344014,
"loss": 3.159,
"step": 39700
},
{
"epoch": 10.69713608957795,
"grad_norm": 0.3488556742668152,
"learning_rate": 0.00043984515955298233,
"loss": 3.1553,
"step": 39750
},
{
"epoch": 10.710594315245478,
"grad_norm": 0.350911408662796,
"learning_rate": 0.00043964319375252453,
"loss": 3.1567,
"step": 39800
},
{
"epoch": 10.724052540913005,
"grad_norm": 0.337378591299057,
"learning_rate": 0.0004394412279520668,
"loss": 3.1561,
"step": 39850
},
{
"epoch": 10.737510766580534,
"grad_norm": 0.3427545726299286,
"learning_rate": 0.000439239262151609,
"loss": 3.1715,
"step": 39900
},
{
"epoch": 10.750968992248062,
"grad_norm": 0.36676934361457825,
"learning_rate": 0.00043903729635115117,
"loss": 3.1629,
"step": 39950
},
{
"epoch": 10.764427217915589,
"grad_norm": 0.38957640528678894,
"learning_rate": 0.00043883533055069337,
"loss": 3.1695,
"step": 40000
},
{
"epoch": 10.764427217915589,
"eval_accuracy": 0.39051004483638435,
"eval_loss": 3.3090879917144775,
"eval_runtime": 53.3592,
"eval_samples_per_second": 337.561,
"eval_steps_per_second": 21.102,
"step": 40000
},
{
"epoch": 10.777885443583118,
"grad_norm": 0.3423351049423218,
"learning_rate": 0.00043863336475023556,
"loss": 3.1569,
"step": 40050
},
{
"epoch": 10.791343669250645,
"grad_norm": 0.3677137494087219,
"learning_rate": 0.0004384313989497778,
"loss": 3.1637,
"step": 40100
},
{
"epoch": 10.804801894918175,
"grad_norm": 0.4070259928703308,
"learning_rate": 0.00043822943314932,
"loss": 3.1542,
"step": 40150
},
{
"epoch": 10.818260120585702,
"grad_norm": 0.35667550563812256,
"learning_rate": 0.0004380274673488622,
"loss": 3.1666,
"step": 40200
},
{
"epoch": 10.83171834625323,
"grad_norm": 0.3461889624595642,
"learning_rate": 0.0004378255015484044,
"loss": 3.1642,
"step": 40250
},
{
"epoch": 10.845176571920758,
"grad_norm": 0.3470584750175476,
"learning_rate": 0.0004376235357479466,
"loss": 3.1641,
"step": 40300
},
{
"epoch": 10.858634797588286,
"grad_norm": 0.3516797423362732,
"learning_rate": 0.00043742156994748885,
"loss": 3.1651,
"step": 40350
},
{
"epoch": 10.872093023255815,
"grad_norm": 0.3457486927509308,
"learning_rate": 0.00043721960414703104,
"loss": 3.1637,
"step": 40400
},
{
"epoch": 10.885551248923342,
"grad_norm": 0.3525344431400299,
"learning_rate": 0.0004370176383465733,
"loss": 3.1693,
"step": 40450
},
{
"epoch": 10.89900947459087,
"grad_norm": 0.3566647469997406,
"learning_rate": 0.0004368156725461155,
"loss": 3.1562,
"step": 40500
},
{
"epoch": 10.912467700258398,
"grad_norm": 0.39564409852027893,
"learning_rate": 0.00043661370674565774,
"loss": 3.163,
"step": 40550
},
{
"epoch": 10.925925925925926,
"grad_norm": 0.35635221004486084,
"learning_rate": 0.00043641174094519994,
"loss": 3.1663,
"step": 40600
},
{
"epoch": 10.939384151593455,
"grad_norm": 0.35757219791412354,
"learning_rate": 0.00043620977514474213,
"loss": 3.1705,
"step": 40650
},
{
"epoch": 10.952842377260982,
"grad_norm": 0.3532245457172394,
"learning_rate": 0.00043600780934428433,
"loss": 3.167,
"step": 40700
},
{
"epoch": 10.96630060292851,
"grad_norm": 0.3705931007862091,
"learning_rate": 0.0004358058435438265,
"loss": 3.1693,
"step": 40750
},
{
"epoch": 10.979758828596038,
"grad_norm": 0.37144991755485535,
"learning_rate": 0.0004356038777433688,
"loss": 3.1758,
"step": 40800
},
{
"epoch": 10.993217054263566,
"grad_norm": 0.3508891761302948,
"learning_rate": 0.00043540191194291097,
"loss": 3.1677,
"step": 40850
},
{
"epoch": 11.006459948320414,
"grad_norm": 0.37817490100860596,
"learning_rate": 0.00043519994614245317,
"loss": 3.1246,
"step": 40900
},
{
"epoch": 11.019918173987941,
"grad_norm": 0.3762407898902893,
"learning_rate": 0.00043499798034199536,
"loss": 3.0683,
"step": 40950
},
{
"epoch": 11.03337639965547,
"grad_norm": 0.3525347411632538,
"learning_rate": 0.00043479601454153756,
"loss": 3.0734,
"step": 41000
},
{
"epoch": 11.03337639965547,
"eval_accuracy": 0.3907509149151707,
"eval_loss": 3.3142662048339844,
"eval_runtime": 53.355,
"eval_samples_per_second": 337.588,
"eval_steps_per_second": 21.104,
"step": 41000
},
{
"epoch": 11.046834625322997,
"grad_norm": 0.36088988184928894,
"learning_rate": 0.0004345940487410798,
"loss": 3.0784,
"step": 41050
},
{
"epoch": 11.060292850990525,
"grad_norm": 0.3609704077243805,
"learning_rate": 0.000434392082940622,
"loss": 3.0779,
"step": 41100
},
{
"epoch": 11.073751076658054,
"grad_norm": 0.35831403732299805,
"learning_rate": 0.0004341901171401642,
"loss": 3.0786,
"step": 41150
},
{
"epoch": 11.087209302325581,
"grad_norm": 0.3548460304737091,
"learning_rate": 0.0004339881513397064,
"loss": 3.0868,
"step": 41200
},
{
"epoch": 11.10066752799311,
"grad_norm": 0.3861493766307831,
"learning_rate": 0.0004337861855392486,
"loss": 3.0904,
"step": 41250
},
{
"epoch": 11.114125753660637,
"grad_norm": 0.4032128155231476,
"learning_rate": 0.00043358421973879084,
"loss": 3.0839,
"step": 41300
},
{
"epoch": 11.127583979328165,
"grad_norm": 0.364173024892807,
"learning_rate": 0.0004333822539383331,
"loss": 3.0909,
"step": 41350
},
{
"epoch": 11.141042204995694,
"grad_norm": 0.35015109181404114,
"learning_rate": 0.0004331802881378753,
"loss": 3.0923,
"step": 41400
},
{
"epoch": 11.154500430663221,
"grad_norm": 0.3518073558807373,
"learning_rate": 0.00043297832233741754,
"loss": 3.0858,
"step": 41450
},
{
"epoch": 11.16795865633075,
"grad_norm": 0.3917633891105652,
"learning_rate": 0.00043277635653695974,
"loss": 3.1072,
"step": 41500
},
{
"epoch": 11.181416881998278,
"grad_norm": 0.3793010711669922,
"learning_rate": 0.00043257439073650193,
"loss": 3.0988,
"step": 41550
},
{
"epoch": 11.194875107665805,
"grad_norm": 0.36705297231674194,
"learning_rate": 0.00043237242493604413,
"loss": 3.0967,
"step": 41600
},
{
"epoch": 11.208333333333334,
"grad_norm": 0.35543742775917053,
"learning_rate": 0.0004321704591355863,
"loss": 3.095,
"step": 41650
},
{
"epoch": 11.221791559000861,
"grad_norm": 0.3734685182571411,
"learning_rate": 0.0004319684933351286,
"loss": 3.1022,
"step": 41700
},
{
"epoch": 11.235249784668389,
"grad_norm": 0.35706183314323425,
"learning_rate": 0.00043176652753467077,
"loss": 3.1099,
"step": 41750
},
{
"epoch": 11.248708010335918,
"grad_norm": 0.4005352556705475,
"learning_rate": 0.00043156456173421297,
"loss": 3.1042,
"step": 41800
},
{
"epoch": 11.262166236003445,
"grad_norm": 0.3816969692707062,
"learning_rate": 0.00043136259593375516,
"loss": 3.1062,
"step": 41850
},
{
"epoch": 11.275624461670974,
"grad_norm": 0.3443428575992584,
"learning_rate": 0.00043116063013329736,
"loss": 3.1235,
"step": 41900
},
{
"epoch": 11.289082687338501,
"grad_norm": 0.3736826479434967,
"learning_rate": 0.0004309586643328396,
"loss": 3.1089,
"step": 41950
},
{
"epoch": 11.302540913006029,
"grad_norm": 0.355075865983963,
"learning_rate": 0.0004307566985323818,
"loss": 3.1179,
"step": 42000
},
{
"epoch": 11.302540913006029,
"eval_accuracy": 0.39087075239685826,
"eval_loss": 3.314624786376953,
"eval_runtime": 53.3716,
"eval_samples_per_second": 337.483,
"eval_steps_per_second": 21.097,
"step": 42000
},
{
"epoch": 11.315999138673558,
"grad_norm": 0.3512454926967621,
"learning_rate": 0.000430554732731924,
"loss": 3.1155,
"step": 42050
},
{
"epoch": 11.329457364341085,
"grad_norm": 0.3447372019290924,
"learning_rate": 0.0004303527669314662,
"loss": 3.1195,
"step": 42100
},
{
"epoch": 11.342915590008614,
"grad_norm": 0.3592742383480072,
"learning_rate": 0.0004301508011310084,
"loss": 3.1171,
"step": 42150
},
{
"epoch": 11.356373815676141,
"grad_norm": 0.39683109521865845,
"learning_rate": 0.00042994883533055065,
"loss": 3.1231,
"step": 42200
},
{
"epoch": 11.369832041343669,
"grad_norm": 0.37733298540115356,
"learning_rate": 0.0004297468695300929,
"loss": 3.1128,
"step": 42250
},
{
"epoch": 11.383290267011198,
"grad_norm": 0.3726079761981964,
"learning_rate": 0.0004295449037296351,
"loss": 3.1142,
"step": 42300
},
{
"epoch": 11.396748492678725,
"grad_norm": 0.3837583363056183,
"learning_rate": 0.0004293429379291773,
"loss": 3.1118,
"step": 42350
},
{
"epoch": 11.410206718346252,
"grad_norm": 0.38088589906692505,
"learning_rate": 0.00042914097212871954,
"loss": 3.1257,
"step": 42400
},
{
"epoch": 11.423664944013781,
"grad_norm": 0.338095098733902,
"learning_rate": 0.00042893900632826173,
"loss": 3.1313,
"step": 42450
},
{
"epoch": 11.437123169681309,
"grad_norm": 0.3455497622489929,
"learning_rate": 0.00042873704052780393,
"loss": 3.1252,
"step": 42500
},
{
"epoch": 11.450581395348838,
"grad_norm": 0.38659998774528503,
"learning_rate": 0.0004285350747273461,
"loss": 3.1153,
"step": 42550
},
{
"epoch": 11.464039621016365,
"grad_norm": 0.34629592299461365,
"learning_rate": 0.0004283331089268883,
"loss": 3.133,
"step": 42600
},
{
"epoch": 11.477497846683892,
"grad_norm": 0.4040473997592926,
"learning_rate": 0.00042813114312643057,
"loss": 3.1331,
"step": 42650
},
{
"epoch": 11.490956072351421,
"grad_norm": 0.3708750307559967,
"learning_rate": 0.00042792917732597277,
"loss": 3.1163,
"step": 42700
},
{
"epoch": 11.504414298018949,
"grad_norm": 0.3524065911769867,
"learning_rate": 0.00042772721152551496,
"loss": 3.1219,
"step": 42750
},
{
"epoch": 11.517872523686478,
"grad_norm": 0.3564377427101135,
"learning_rate": 0.00042752524572505716,
"loss": 3.1337,
"step": 42800
},
{
"epoch": 11.531330749354005,
"grad_norm": 0.3928210437297821,
"learning_rate": 0.00042732327992459936,
"loss": 3.1308,
"step": 42850
},
{
"epoch": 11.544788975021532,
"grad_norm": 0.3760032653808594,
"learning_rate": 0.0004271213141241416,
"loss": 3.1333,
"step": 42900
},
{
"epoch": 11.558247200689062,
"grad_norm": 0.3637098968029022,
"learning_rate": 0.0004269193483236838,
"loss": 3.1287,
"step": 42950
},
{
"epoch": 11.571705426356589,
"grad_norm": 0.3505527377128601,
"learning_rate": 0.000426717382523226,
"loss": 3.1317,
"step": 43000
},
{
"epoch": 11.571705426356589,
"eval_accuracy": 0.39133652146666303,
"eval_loss": 3.3082401752471924,
"eval_runtime": 53.3948,
"eval_samples_per_second": 337.336,
"eval_steps_per_second": 21.088,
"step": 43000
},
{
"epoch": 11.585163652024118,
"grad_norm": 0.36543312668800354,
"learning_rate": 0.0004265154167227682,
"loss": 3.1412,
"step": 43050
},
{
"epoch": 11.598621877691645,
"grad_norm": 0.3579201102256775,
"learning_rate": 0.0004263134509223105,
"loss": 3.1364,
"step": 43100
},
{
"epoch": 11.612080103359173,
"grad_norm": 0.385629802942276,
"learning_rate": 0.0004261114851218527,
"loss": 3.1325,
"step": 43150
},
{
"epoch": 11.625538329026702,
"grad_norm": 0.3662668764591217,
"learning_rate": 0.0004259095193213949,
"loss": 3.1262,
"step": 43200
},
{
"epoch": 11.638996554694229,
"grad_norm": 0.38772115111351013,
"learning_rate": 0.0004257075535209371,
"loss": 3.132,
"step": 43250
},
{
"epoch": 11.652454780361758,
"grad_norm": 0.356647789478302,
"learning_rate": 0.00042550558772047934,
"loss": 3.137,
"step": 43300
},
{
"epoch": 11.665913006029285,
"grad_norm": 0.35813525319099426,
"learning_rate": 0.00042530362192002153,
"loss": 3.1488,
"step": 43350
},
{
"epoch": 11.679371231696813,
"grad_norm": 0.3788353204727173,
"learning_rate": 0.00042510165611956373,
"loss": 3.1322,
"step": 43400
},
{
"epoch": 11.692829457364342,
"grad_norm": 0.33950915932655334,
"learning_rate": 0.0004248996903191059,
"loss": 3.1315,
"step": 43450
},
{
"epoch": 11.706287683031869,
"grad_norm": 0.3577655255794525,
"learning_rate": 0.0004246977245186481,
"loss": 3.1414,
"step": 43500
},
{
"epoch": 11.719745908699396,
"grad_norm": 0.36303114891052246,
"learning_rate": 0.0004244957587181904,
"loss": 3.1379,
"step": 43550
},
{
"epoch": 11.733204134366925,
"grad_norm": 0.3299844563007355,
"learning_rate": 0.00042429379291773257,
"loss": 3.1317,
"step": 43600
},
{
"epoch": 11.746662360034453,
"grad_norm": 0.36776620149612427,
"learning_rate": 0.00042409182711727477,
"loss": 3.1367,
"step": 43650
},
{
"epoch": 11.760120585701982,
"grad_norm": 0.36119797825813293,
"learning_rate": 0.00042388986131681696,
"loss": 3.1506,
"step": 43700
},
{
"epoch": 11.773578811369509,
"grad_norm": 0.3472893536090851,
"learning_rate": 0.00042368789551635916,
"loss": 3.1361,
"step": 43750
},
{
"epoch": 11.787037037037036,
"grad_norm": 0.3448840081691742,
"learning_rate": 0.0004234859297159014,
"loss": 3.1447,
"step": 43800
},
{
"epoch": 11.800495262704565,
"grad_norm": 0.3472510278224945,
"learning_rate": 0.0004232839639154436,
"loss": 3.1501,
"step": 43850
},
{
"epoch": 11.813953488372093,
"grad_norm": 0.3915238678455353,
"learning_rate": 0.0004230819981149858,
"loss": 3.1552,
"step": 43900
},
{
"epoch": 11.827411714039622,
"grad_norm": 0.3639262914657593,
"learning_rate": 0.000422880032314528,
"loss": 3.1447,
"step": 43950
},
{
"epoch": 11.840869939707149,
"grad_norm": 0.3748035430908203,
"learning_rate": 0.0004226780665140703,
"loss": 3.1587,
"step": 44000
},
{
"epoch": 11.840869939707149,
"eval_accuracy": 0.39165963721511166,
"eval_loss": 3.3021323680877686,
"eval_runtime": 53.4241,
"eval_samples_per_second": 337.152,
"eval_steps_per_second": 21.077,
"step": 44000
},
{
"epoch": 11.854328165374676,
"grad_norm": 0.36416128277778625,
"learning_rate": 0.0004224761007136125,
"loss": 3.138,
"step": 44050
},
{
"epoch": 11.867786391042205,
"grad_norm": 0.3731189966201782,
"learning_rate": 0.0004222741349131547,
"loss": 3.142,
"step": 44100
},
{
"epoch": 11.881244616709733,
"grad_norm": 0.35165873169898987,
"learning_rate": 0.0004220721691126969,
"loss": 3.1496,
"step": 44150
},
{
"epoch": 11.89470284237726,
"grad_norm": 0.36347705125808716,
"learning_rate": 0.0004218702033122391,
"loss": 3.1334,
"step": 44200
},
{
"epoch": 11.90816106804479,
"grad_norm": 0.3625389337539673,
"learning_rate": 0.00042166823751178134,
"loss": 3.1409,
"step": 44250
},
{
"epoch": 11.921619293712316,
"grad_norm": 0.3359973132610321,
"learning_rate": 0.00042146627171132353,
"loss": 3.1419,
"step": 44300
},
{
"epoch": 11.935077519379846,
"grad_norm": 0.3568677604198456,
"learning_rate": 0.00042126430591086573,
"loss": 3.1484,
"step": 44350
},
{
"epoch": 11.948535745047373,
"grad_norm": 0.37474843859672546,
"learning_rate": 0.0004210623401104079,
"loss": 3.1633,
"step": 44400
},
{
"epoch": 11.9619939707149,
"grad_norm": 0.37014254927635193,
"learning_rate": 0.0004208603743099501,
"loss": 3.1547,
"step": 44450
},
{
"epoch": 11.97545219638243,
"grad_norm": 0.3610021471977234,
"learning_rate": 0.00042065840850949237,
"loss": 3.1397,
"step": 44500
},
{
"epoch": 11.988910422049956,
"grad_norm": 0.36313939094543457,
"learning_rate": 0.00042045644270903457,
"loss": 3.1523,
"step": 44550
},
{
"epoch": 12.002153316106805,
"grad_norm": 0.3740599751472473,
"learning_rate": 0.00042025447690857676,
"loss": 3.1279,
"step": 44600
},
{
"epoch": 12.015611541774332,
"grad_norm": 0.3721098303794861,
"learning_rate": 0.00042005251110811896,
"loss": 3.0501,
"step": 44650
},
{
"epoch": 12.029069767441861,
"grad_norm": 0.37249475717544556,
"learning_rate": 0.00041985054530766115,
"loss": 3.0593,
"step": 44700
},
{
"epoch": 12.042527993109388,
"grad_norm": 0.37305739521980286,
"learning_rate": 0.0004196485795072034,
"loss": 3.0575,
"step": 44750
},
{
"epoch": 12.055986218776916,
"grad_norm": 0.3568457365036011,
"learning_rate": 0.0004194466137067456,
"loss": 3.0538,
"step": 44800
},
{
"epoch": 12.069444444444445,
"grad_norm": 0.3669044077396393,
"learning_rate": 0.0004192446479062878,
"loss": 3.0684,
"step": 44850
},
{
"epoch": 12.082902670111972,
"grad_norm": 0.37108471989631653,
"learning_rate": 0.00041904268210583005,
"loss": 3.068,
"step": 44900
},
{
"epoch": 12.096360895779501,
"grad_norm": 0.37075528502464294,
"learning_rate": 0.0004188407163053723,
"loss": 3.0668,
"step": 44950
},
{
"epoch": 12.109819121447028,
"grad_norm": 0.3636740446090698,
"learning_rate": 0.0004186387505049145,
"loss": 3.0608,
"step": 45000
},
{
"epoch": 12.109819121447028,
"eval_accuracy": 0.39137433057239945,
"eval_loss": 3.3102362155914307,
"eval_runtime": 53.3835,
"eval_samples_per_second": 337.408,
"eval_steps_per_second": 21.093,
"step": 45000
},
{
"epoch": 12.123277347114556,
"grad_norm": 0.3946690857410431,
"learning_rate": 0.0004184367847044567,
"loss": 3.069,
"step": 45050
},
{
"epoch": 12.136735572782085,
"grad_norm": 0.4067884683609009,
"learning_rate": 0.0004182348189039989,
"loss": 3.0741,
"step": 45100
},
{
"epoch": 12.150193798449612,
"grad_norm": 0.3845987021923065,
"learning_rate": 0.00041803285310354114,
"loss": 3.0809,
"step": 45150
},
{
"epoch": 12.163652024117141,
"grad_norm": 0.3612961769104004,
"learning_rate": 0.00041783088730308333,
"loss": 3.0856,
"step": 45200
},
{
"epoch": 12.177110249784668,
"grad_norm": 0.32780522108078003,
"learning_rate": 0.00041762892150262553,
"loss": 3.0784,
"step": 45250
},
{
"epoch": 12.190568475452196,
"grad_norm": 0.3584006130695343,
"learning_rate": 0.0004174269557021677,
"loss": 3.0792,
"step": 45300
},
{
"epoch": 12.204026701119725,
"grad_norm": 0.3582470715045929,
"learning_rate": 0.0004172249899017099,
"loss": 3.0836,
"step": 45350
},
{
"epoch": 12.217484926787252,
"grad_norm": 0.36709001660346985,
"learning_rate": 0.00041702302410125217,
"loss": 3.0843,
"step": 45400
},
{
"epoch": 12.230943152454781,
"grad_norm": 0.3590288758277893,
"learning_rate": 0.00041682105830079437,
"loss": 3.0847,
"step": 45450
},
{
"epoch": 12.244401378122308,
"grad_norm": 0.3631494641304016,
"learning_rate": 0.00041661909250033656,
"loss": 3.0913,
"step": 45500
},
{
"epoch": 12.257859603789836,
"grad_norm": 0.3625164031982422,
"learning_rate": 0.00041641712669987876,
"loss": 3.0862,
"step": 45550
},
{
"epoch": 12.271317829457365,
"grad_norm": 0.39201679825782776,
"learning_rate": 0.00041621516089942096,
"loss": 3.0965,
"step": 45600
},
{
"epoch": 12.284776055124892,
"grad_norm": 0.38754960894584656,
"learning_rate": 0.0004160131950989632,
"loss": 3.0899,
"step": 45650
},
{
"epoch": 12.298234280792421,
"grad_norm": 0.36467859148979187,
"learning_rate": 0.0004158112292985054,
"loss": 3.102,
"step": 45700
},
{
"epoch": 12.311692506459949,
"grad_norm": 0.38860246539115906,
"learning_rate": 0.0004156092634980476,
"loss": 3.094,
"step": 45750
},
{
"epoch": 12.325150732127476,
"grad_norm": 0.36453795433044434,
"learning_rate": 0.00041540729769758985,
"loss": 3.0931,
"step": 45800
},
{
"epoch": 12.338608957795005,
"grad_norm": 0.39302724599838257,
"learning_rate": 0.0004152053318971321,
"loss": 3.1034,
"step": 45850
},
{
"epoch": 12.352067183462532,
"grad_norm": 0.38808274269104004,
"learning_rate": 0.0004150033660966743,
"loss": 3.0973,
"step": 45900
},
{
"epoch": 12.36552540913006,
"grad_norm": 0.3711463212966919,
"learning_rate": 0.0004148014002962165,
"loss": 3.1053,
"step": 45950
},
{
"epoch": 12.378983634797589,
"grad_norm": 0.3651505708694458,
"learning_rate": 0.0004145994344957587,
"loss": 3.0961,
"step": 46000
},
{
"epoch": 12.378983634797589,
"eval_accuracy": 0.3916768034182908,
"eval_loss": 3.307568073272705,
"eval_runtime": 53.4044,
"eval_samples_per_second": 337.275,
"eval_steps_per_second": 21.084,
"step": 46000
},
{
"epoch": 12.392441860465116,
"grad_norm": 0.3489798903465271,
"learning_rate": 0.0004143974686953009,
"loss": 3.1111,
"step": 46050
},
{
"epoch": 12.405900086132645,
"grad_norm": 0.37551695108413696,
"learning_rate": 0.00041419550289484313,
"loss": 3.1072,
"step": 46100
},
{
"epoch": 12.419358311800172,
"grad_norm": 0.35827645659446716,
"learning_rate": 0.00041399353709438533,
"loss": 3.1013,
"step": 46150
},
{
"epoch": 12.4328165374677,
"grad_norm": 0.3637438416481018,
"learning_rate": 0.0004137915712939275,
"loss": 3.1106,
"step": 46200
},
{
"epoch": 12.446274763135229,
"grad_norm": 0.3819226920604706,
"learning_rate": 0.0004135896054934697,
"loss": 3.1055,
"step": 46250
},
{
"epoch": 12.459732988802756,
"grad_norm": 0.36781030893325806,
"learning_rate": 0.0004133876396930119,
"loss": 3.1078,
"step": 46300
},
{
"epoch": 12.473191214470285,
"grad_norm": 0.39347222447395325,
"learning_rate": 0.00041318567389255417,
"loss": 3.1151,
"step": 46350
},
{
"epoch": 12.486649440137812,
"grad_norm": 0.39293286204338074,
"learning_rate": 0.00041298370809209636,
"loss": 3.1106,
"step": 46400
},
{
"epoch": 12.50010766580534,
"grad_norm": 0.37734219431877136,
"learning_rate": 0.00041278174229163856,
"loss": 3.106,
"step": 46450
},
{
"epoch": 12.513565891472869,
"grad_norm": 0.3757116198539734,
"learning_rate": 0.00041257977649118076,
"loss": 3.1061,
"step": 46500
},
{
"epoch": 12.527024117140396,
"grad_norm": 0.4108484089374542,
"learning_rate": 0.00041237781069072295,
"loss": 3.1093,
"step": 46550
},
{
"epoch": 12.540482342807923,
"grad_norm": 0.3842467963695526,
"learning_rate": 0.0004121758448902652,
"loss": 3.1056,
"step": 46600
},
{
"epoch": 12.553940568475452,
"grad_norm": 0.38736289739608765,
"learning_rate": 0.0004119738790898074,
"loss": 3.1208,
"step": 46650
},
{
"epoch": 12.56739879414298,
"grad_norm": 0.3820643126964569,
"learning_rate": 0.00041177191328934965,
"loss": 3.1143,
"step": 46700
},
{
"epoch": 12.580857019810509,
"grad_norm": 0.37509894371032715,
"learning_rate": 0.00041156994748889185,
"loss": 3.1174,
"step": 46750
},
{
"epoch": 12.594315245478036,
"grad_norm": 0.35530561208724976,
"learning_rate": 0.0004113679816884341,
"loss": 3.1197,
"step": 46800
},
{
"epoch": 12.607773471145563,
"grad_norm": 0.3885257840156555,
"learning_rate": 0.0004111660158879763,
"loss": 3.1136,
"step": 46850
},
{
"epoch": 12.621231696813092,
"grad_norm": 0.38516077399253845,
"learning_rate": 0.0004109640500875185,
"loss": 3.1234,
"step": 46900
},
{
"epoch": 12.63468992248062,
"grad_norm": 0.3831053674221039,
"learning_rate": 0.0004107620842870607,
"loss": 3.1207,
"step": 46950
},
{
"epoch": 12.648148148148149,
"grad_norm": 0.3733752965927124,
"learning_rate": 0.0004105601184866029,
"loss": 3.1092,
"step": 47000
},
{
"epoch": 12.648148148148149,
"eval_accuracy": 0.3917694791860873,
"eval_loss": 3.3039400577545166,
"eval_runtime": 53.5807,
"eval_samples_per_second": 336.166,
"eval_steps_per_second": 21.015,
"step": 47000
},
{
"epoch": 12.661606373815676,
"grad_norm": 0.38987475633621216,
"learning_rate": 0.00041035815268614513,
"loss": 3.114,
"step": 47050
},
{
"epoch": 12.675064599483203,
"grad_norm": 0.40705275535583496,
"learning_rate": 0.0004101561868856873,
"loss": 3.1164,
"step": 47100
},
{
"epoch": 12.688522825150732,
"grad_norm": 0.373837411403656,
"learning_rate": 0.0004099542210852295,
"loss": 3.1208,
"step": 47150
},
{
"epoch": 12.70198105081826,
"grad_norm": 0.3675236701965332,
"learning_rate": 0.0004097522552847717,
"loss": 3.1212,
"step": 47200
},
{
"epoch": 12.715439276485789,
"grad_norm": 0.35344648361206055,
"learning_rate": 0.00040955028948431397,
"loss": 3.1104,
"step": 47250
},
{
"epoch": 12.728897502153316,
"grad_norm": 0.3793870508670807,
"learning_rate": 0.00040934832368385616,
"loss": 3.1225,
"step": 47300
},
{
"epoch": 12.742355727820843,
"grad_norm": 0.4043644666671753,
"learning_rate": 0.00040914635788339836,
"loss": 3.1305,
"step": 47350
},
{
"epoch": 12.755813953488373,
"grad_norm": 0.37010902166366577,
"learning_rate": 0.00040894439208294056,
"loss": 3.1237,
"step": 47400
},
{
"epoch": 12.7692721791559,
"grad_norm": 0.4060969054698944,
"learning_rate": 0.00040874242628248275,
"loss": 3.1249,
"step": 47450
},
{
"epoch": 12.782730404823429,
"grad_norm": 0.3517776131629944,
"learning_rate": 0.000408540460482025,
"loss": 3.1302,
"step": 47500
},
{
"epoch": 12.796188630490956,
"grad_norm": 0.34859320521354675,
"learning_rate": 0.00040833849468156725,
"loss": 3.1223,
"step": 47550
},
{
"epoch": 12.809646856158484,
"grad_norm": 0.3657214641571045,
"learning_rate": 0.00040813652888110945,
"loss": 3.1296,
"step": 47600
},
{
"epoch": 12.823105081826013,
"grad_norm": 0.3647889494895935,
"learning_rate": 0.00040793456308065165,
"loss": 3.1298,
"step": 47650
},
{
"epoch": 12.83656330749354,
"grad_norm": 0.37984415888786316,
"learning_rate": 0.0004077325972801939,
"loss": 3.1326,
"step": 47700
},
{
"epoch": 12.850021533161069,
"grad_norm": 0.3746028244495392,
"learning_rate": 0.0004075306314797361,
"loss": 3.1221,
"step": 47750
},
{
"epoch": 12.863479758828596,
"grad_norm": 0.360163152217865,
"learning_rate": 0.0004073286656792783,
"loss": 3.1311,
"step": 47800
},
{
"epoch": 12.876937984496124,
"grad_norm": 0.4016077518463135,
"learning_rate": 0.0004071266998788205,
"loss": 3.1373,
"step": 47850
},
{
"epoch": 12.890396210163653,
"grad_norm": 0.3723677694797516,
"learning_rate": 0.0004069247340783627,
"loss": 3.1192,
"step": 47900
},
{
"epoch": 12.90385443583118,
"grad_norm": 0.3679752051830292,
"learning_rate": 0.00040672276827790493,
"loss": 3.1305,
"step": 47950
},
{
"epoch": 12.917312661498707,
"grad_norm": 0.3905462920665741,
"learning_rate": 0.0004065208024774471,
"loss": 3.129,
"step": 48000
},
{
"epoch": 12.917312661498707,
"eval_accuracy": 0.3922937002641857,
"eval_loss": 3.2970693111419678,
"eval_runtime": 53.7542,
"eval_samples_per_second": 335.081,
"eval_steps_per_second": 20.947,
"step": 48000
},
{
"epoch": 12.930770887166236,
"grad_norm": 0.36770474910736084,
"learning_rate": 0.0004063188366769893,
"loss": 3.1407,
"step": 48050
},
{
"epoch": 12.944229112833764,
"grad_norm": 0.3561168611049652,
"learning_rate": 0.0004061168708765315,
"loss": 3.1399,
"step": 48100
},
{
"epoch": 12.957687338501293,
"grad_norm": 0.3862304389476776,
"learning_rate": 0.0004059149050760737,
"loss": 3.1351,
"step": 48150
},
{
"epoch": 12.97114556416882,
"grad_norm": 0.36406269669532776,
"learning_rate": 0.00040571293927561597,
"loss": 3.1314,
"step": 48200
},
{
"epoch": 12.984603789836347,
"grad_norm": 0.3505527377128601,
"learning_rate": 0.00040551097347515816,
"loss": 3.129,
"step": 48250
},
{
"epoch": 12.998062015503876,
"grad_norm": 0.37477901577949524,
"learning_rate": 0.00040530900767470036,
"loss": 3.1402,
"step": 48300
},
{
"epoch": 13.011304909560723,
"grad_norm": 0.39581868052482605,
"learning_rate": 0.00040510704187424255,
"loss": 3.0524,
"step": 48350
},
{
"epoch": 13.024763135228252,
"grad_norm": 0.35605454444885254,
"learning_rate": 0.00040490507607378475,
"loss": 3.0409,
"step": 48400
},
{
"epoch": 13.038221360895779,
"grad_norm": 0.36419981718063354,
"learning_rate": 0.00040470311027332705,
"loss": 3.0414,
"step": 48450
},
{
"epoch": 13.051679586563308,
"grad_norm": 0.377587229013443,
"learning_rate": 0.00040450114447286925,
"loss": 3.043,
"step": 48500
},
{
"epoch": 13.065137812230835,
"grad_norm": 0.3738135099411011,
"learning_rate": 0.00040429917867241145,
"loss": 3.0349,
"step": 48550
},
{
"epoch": 13.078596037898363,
"grad_norm": 0.37447571754455566,
"learning_rate": 0.00040409721287195364,
"loss": 3.0418,
"step": 48600
},
{
"epoch": 13.092054263565892,
"grad_norm": 0.37799814343452454,
"learning_rate": 0.0004038952470714959,
"loss": 3.0625,
"step": 48650
},
{
"epoch": 13.10551248923342,
"grad_norm": 0.36537399888038635,
"learning_rate": 0.0004036932812710381,
"loss": 3.0552,
"step": 48700
},
{
"epoch": 13.118970714900948,
"grad_norm": 0.3800034821033478,
"learning_rate": 0.0004034913154705803,
"loss": 3.0689,
"step": 48750
},
{
"epoch": 13.132428940568476,
"grad_norm": 0.36928868293762207,
"learning_rate": 0.0004032893496701225,
"loss": 3.059,
"step": 48800
},
{
"epoch": 13.145887166236003,
"grad_norm": 0.36283737421035767,
"learning_rate": 0.0004030873838696647,
"loss": 3.066,
"step": 48850
},
{
"epoch": 13.159345391903532,
"grad_norm": 0.3781079053878784,
"learning_rate": 0.00040288541806920693,
"loss": 3.0596,
"step": 48900
},
{
"epoch": 13.17280361757106,
"grad_norm": 0.37426838278770447,
"learning_rate": 0.0004026834522687491,
"loss": 3.0596,
"step": 48950
},
{
"epoch": 13.186261843238588,
"grad_norm": 0.36410465836524963,
"learning_rate": 0.0004024814864682913,
"loss": 3.0675,
"step": 49000
},
{
"epoch": 13.186261843238588,
"eval_accuracy": 0.3923700790036475,
"eval_loss": 3.3071610927581787,
"eval_runtime": 53.2206,
"eval_samples_per_second": 338.441,
"eval_steps_per_second": 21.157,
"step": 49000
},
{
"epoch": 13.199720068906116,
"grad_norm": 0.39360588788986206,
"learning_rate": 0.0004022795206678335,
"loss": 3.0699,
"step": 49050
},
{
"epoch": 13.213178294573643,
"grad_norm": 0.3912392556667328,
"learning_rate": 0.00040207755486737577,
"loss": 3.0677,
"step": 49100
},
{
"epoch": 13.226636520241172,
"grad_norm": 0.403340607881546,
"learning_rate": 0.00040187558906691796,
"loss": 3.061,
"step": 49150
},
{
"epoch": 13.2400947459087,
"grad_norm": 0.3977563977241516,
"learning_rate": 0.00040167362326646016,
"loss": 3.0714,
"step": 49200
},
{
"epoch": 13.253552971576227,
"grad_norm": 0.44516828656196594,
"learning_rate": 0.00040147165746600235,
"loss": 3.0773,
"step": 49250
},
{
"epoch": 13.267011197243756,
"grad_norm": 0.38486993312835693,
"learning_rate": 0.00040126969166554455,
"loss": 3.0755,
"step": 49300
},
{
"epoch": 13.280469422911283,
"grad_norm": 0.37119653820991516,
"learning_rate": 0.00040106772586508686,
"loss": 3.0801,
"step": 49350
},
{
"epoch": 13.293927648578812,
"grad_norm": 0.37240782380104065,
"learning_rate": 0.00040086576006462905,
"loss": 3.0811,
"step": 49400
},
{
"epoch": 13.30738587424634,
"grad_norm": 0.362568199634552,
"learning_rate": 0.00040066379426417125,
"loss": 3.0827,
"step": 49450
},
{
"epoch": 13.320844099913867,
"grad_norm": 0.3640483617782593,
"learning_rate": 0.00040046182846371344,
"loss": 3.0916,
"step": 49500
},
{
"epoch": 13.334302325581396,
"grad_norm": 0.37040454149246216,
"learning_rate": 0.0004002598626632557,
"loss": 3.0813,
"step": 49550
},
{
"epoch": 13.347760551248923,
"grad_norm": 0.38464635610580444,
"learning_rate": 0.0004000578968627979,
"loss": 3.099,
"step": 49600
},
{
"epoch": 13.361218776916452,
"grad_norm": 0.38465845584869385,
"learning_rate": 0.0003998559310623401,
"loss": 3.0855,
"step": 49650
},
{
"epoch": 13.37467700258398,
"grad_norm": 0.3531036376953125,
"learning_rate": 0.0003996539652618823,
"loss": 3.0837,
"step": 49700
},
{
"epoch": 13.388135228251507,
"grad_norm": 0.38886111974716187,
"learning_rate": 0.0003994519994614245,
"loss": 3.0925,
"step": 49750
},
{
"epoch": 13.401593453919036,
"grad_norm": 0.3840842843055725,
"learning_rate": 0.00039925003366096673,
"loss": 3.0875,
"step": 49800
},
{
"epoch": 13.415051679586563,
"grad_norm": 0.3760302662849426,
"learning_rate": 0.0003990480678605089,
"loss": 3.0873,
"step": 49850
},
{
"epoch": 13.428509905254092,
"grad_norm": 0.36125680804252625,
"learning_rate": 0.0003988461020600511,
"loss": 3.0911,
"step": 49900
},
{
"epoch": 13.44196813092162,
"grad_norm": 0.38114649057388306,
"learning_rate": 0.0003986441362595933,
"loss": 3.0869,
"step": 49950
},
{
"epoch": 13.455426356589147,
"grad_norm": 0.38252297043800354,
"learning_rate": 0.0003984421704591355,
"loss": 3.0914,
"step": 50000
},
{
"epoch": 13.455426356589147,
"eval_accuracy": 0.3924209257320517,
"eval_loss": 3.3010644912719727,
"eval_runtime": 53.2801,
"eval_samples_per_second": 338.062,
"eval_steps_per_second": 21.134,
"step": 50000
},
{
"epoch": 13.468884582256676,
"grad_norm": 0.3857160806655884,
"learning_rate": 0.00039824020465867776,
"loss": 3.0851,
"step": 50050
},
{
"epoch": 13.482342807924203,
"grad_norm": 0.37299609184265137,
"learning_rate": 0.00039803823885821996,
"loss": 3.0893,
"step": 50100
},
{
"epoch": 13.49580103359173,
"grad_norm": 0.350725382566452,
"learning_rate": 0.00039783627305776216,
"loss": 3.0979,
"step": 50150
},
{
"epoch": 13.50925925925926,
"grad_norm": 0.37564557790756226,
"learning_rate": 0.00039763430725730435,
"loss": 3.0976,
"step": 50200
},
{
"epoch": 13.522717484926787,
"grad_norm": 0.3750132918357849,
"learning_rate": 0.00039743234145684666,
"loss": 3.1028,
"step": 50250
},
{
"epoch": 13.536175710594316,
"grad_norm": 0.38835328817367554,
"learning_rate": 0.00039723037565638885,
"loss": 3.0929,
"step": 50300
},
{
"epoch": 13.549633936261843,
"grad_norm": 0.38070812821388245,
"learning_rate": 0.00039702840985593105,
"loss": 3.0919,
"step": 50350
},
{
"epoch": 13.56309216192937,
"grad_norm": 0.393497496843338,
"learning_rate": 0.00039682644405547324,
"loss": 3.0975,
"step": 50400
},
{
"epoch": 13.5765503875969,
"grad_norm": 0.3958127200603485,
"learning_rate": 0.00039662447825501544,
"loss": 3.0977,
"step": 50450
},
{
"epoch": 13.590008613264427,
"grad_norm": 0.4099980592727661,
"learning_rate": 0.0003964225124545577,
"loss": 3.1074,
"step": 50500
},
{
"epoch": 13.603466838931956,
"grad_norm": 0.37544602155685425,
"learning_rate": 0.0003962205466540999,
"loss": 3.1021,
"step": 50550
},
{
"epoch": 13.616925064599483,
"grad_norm": 0.3651670515537262,
"learning_rate": 0.0003960185808536421,
"loss": 3.1032,
"step": 50600
},
{
"epoch": 13.63038329026701,
"grad_norm": 0.4086431562900543,
"learning_rate": 0.0003958166150531843,
"loss": 3.0982,
"step": 50650
},
{
"epoch": 13.64384151593454,
"grad_norm": 0.37201717495918274,
"learning_rate": 0.0003956146492527265,
"loss": 3.1072,
"step": 50700
},
{
"epoch": 13.657299741602067,
"grad_norm": 0.372783899307251,
"learning_rate": 0.0003954126834522687,
"loss": 3.1025,
"step": 50750
},
{
"epoch": 13.670757967269594,
"grad_norm": 0.38944682478904724,
"learning_rate": 0.0003952107176518109,
"loss": 3.101,
"step": 50800
},
{
"epoch": 13.684216192937123,
"grad_norm": 0.3839119076728821,
"learning_rate": 0.0003950087518513531,
"loss": 3.1006,
"step": 50850
},
{
"epoch": 13.69767441860465,
"grad_norm": 0.38758784532546997,
"learning_rate": 0.0003948067860508953,
"loss": 3.0906,
"step": 50900
},
{
"epoch": 13.71113264427218,
"grad_norm": 0.3955155611038208,
"learning_rate": 0.0003946048202504375,
"loss": 3.1069,
"step": 50950
},
{
"epoch": 13.724590869939707,
"grad_norm": 0.3915644586086273,
"learning_rate": 0.00039440285444997976,
"loss": 3.1114,
"step": 51000
},
{
"epoch": 13.724590869939707,
"eval_accuracy": 0.3928451917030308,
"eval_loss": 3.2942299842834473,
"eval_runtime": 53.1633,
"eval_samples_per_second": 338.805,
"eval_steps_per_second": 21.18,
"step": 51000
},
{
"epoch": 13.738049095607234,
"grad_norm": 0.3906634449958801,
"learning_rate": 0.00039420088864952196,
"loss": 3.1124,
"step": 51050
},
{
"epoch": 13.751507321274763,
"grad_norm": 0.3947088420391083,
"learning_rate": 0.00039399892284906415,
"loss": 3.1152,
"step": 51100
},
{
"epoch": 13.76496554694229,
"grad_norm": 0.37320712208747864,
"learning_rate": 0.00039379695704860646,
"loss": 3.1083,
"step": 51150
},
{
"epoch": 13.77842377260982,
"grad_norm": 0.3770774006843567,
"learning_rate": 0.00039359499124814865,
"loss": 3.1003,
"step": 51200
},
{
"epoch": 13.791881998277347,
"grad_norm": 0.35937607288360596,
"learning_rate": 0.00039339302544769085,
"loss": 3.1084,
"step": 51250
},
{
"epoch": 13.805340223944874,
"grad_norm": 0.3897993564605713,
"learning_rate": 0.00039319105964723305,
"loss": 3.0999,
"step": 51300
},
{
"epoch": 13.818798449612403,
"grad_norm": 0.3844742178916931,
"learning_rate": 0.00039298909384677524,
"loss": 3.1092,
"step": 51350
},
{
"epoch": 13.83225667527993,
"grad_norm": 0.3903350532054901,
"learning_rate": 0.0003927871280463175,
"loss": 3.1111,
"step": 51400
},
{
"epoch": 13.84571490094746,
"grad_norm": 0.3959101438522339,
"learning_rate": 0.0003925851622458597,
"loss": 3.1099,
"step": 51450
},
{
"epoch": 13.859173126614987,
"grad_norm": 0.3868245482444763,
"learning_rate": 0.0003923831964454019,
"loss": 3.1137,
"step": 51500
},
{
"epoch": 13.872631352282514,
"grad_norm": 0.3856070935726166,
"learning_rate": 0.0003921812306449441,
"loss": 3.1112,
"step": 51550
},
{
"epoch": 13.886089577950044,
"grad_norm": 0.38117513060569763,
"learning_rate": 0.0003919792648444863,
"loss": 3.1144,
"step": 51600
},
{
"epoch": 13.89954780361757,
"grad_norm": 0.34406477212905884,
"learning_rate": 0.0003917772990440285,
"loss": 3.1228,
"step": 51650
},
{
"epoch": 13.9130060292851,
"grad_norm": 0.3809207081794739,
"learning_rate": 0.0003915753332435707,
"loss": 3.1175,
"step": 51700
},
{
"epoch": 13.926464254952627,
"grad_norm": 0.38697007298469543,
"learning_rate": 0.0003913733674431129,
"loss": 3.1133,
"step": 51750
},
{
"epoch": 13.939922480620154,
"grad_norm": 0.37401506304740906,
"learning_rate": 0.0003911714016426551,
"loss": 3.1094,
"step": 51800
},
{
"epoch": 13.953380706287684,
"grad_norm": 0.385299950838089,
"learning_rate": 0.0003909694358421973,
"loss": 3.1148,
"step": 51850
},
{
"epoch": 13.96683893195521,
"grad_norm": 0.3646024167537689,
"learning_rate": 0.00039076747004173956,
"loss": 3.1251,
"step": 51900
},
{
"epoch": 13.98029715762274,
"grad_norm": 0.39629772305488586,
"learning_rate": 0.00039056550424128176,
"loss": 3.1245,
"step": 51950
},
{
"epoch": 13.993755383290267,
"grad_norm": 0.3806076645851135,
"learning_rate": 0.00039036353844082395,
"loss": 3.1143,
"step": 52000
},
{
"epoch": 13.993755383290267,
"eval_accuracy": 0.3933573529801615,
"eval_loss": 3.2886648178100586,
"eval_runtime": 53.2025,
"eval_samples_per_second": 338.556,
"eval_steps_per_second": 21.164,
"step": 52000
},
{
"epoch": 14.006998277347115,
"grad_norm": 0.39359694719314575,
"learning_rate": 0.0003901615726403662,
"loss": 3.0789,
"step": 52050
},
{
"epoch": 14.020456503014643,
"grad_norm": 0.36778563261032104,
"learning_rate": 0.00038995960683990845,
"loss": 3.0137,
"step": 52100
},
{
"epoch": 14.03391472868217,
"grad_norm": 0.3619430959224701,
"learning_rate": 0.00038975764103945065,
"loss": 3.0305,
"step": 52150
},
{
"epoch": 14.047372954349699,
"grad_norm": 0.39204686880111694,
"learning_rate": 0.00038955567523899285,
"loss": 3.0243,
"step": 52200
},
{
"epoch": 14.060831180017226,
"grad_norm": 0.37758973240852356,
"learning_rate": 0.00038935370943853504,
"loss": 3.0231,
"step": 52250
},
{
"epoch": 14.074289405684755,
"grad_norm": 0.38672971725463867,
"learning_rate": 0.00038915174363807724,
"loss": 3.0347,
"step": 52300
},
{
"epoch": 14.087747631352283,
"grad_norm": 0.3783572018146515,
"learning_rate": 0.0003889497778376195,
"loss": 3.0336,
"step": 52350
},
{
"epoch": 14.10120585701981,
"grad_norm": 0.3870309293270111,
"learning_rate": 0.0003887478120371617,
"loss": 3.0386,
"step": 52400
},
{
"epoch": 14.114664082687339,
"grad_norm": 0.39776405692100525,
"learning_rate": 0.0003885458462367039,
"loss": 3.027,
"step": 52450
},
{
"epoch": 14.128122308354866,
"grad_norm": 0.44890081882476807,
"learning_rate": 0.0003883438804362461,
"loss": 3.0538,
"step": 52500
},
{
"epoch": 14.141580534022394,
"grad_norm": 0.37647005915641785,
"learning_rate": 0.0003881419146357883,
"loss": 3.0411,
"step": 52550
},
{
"epoch": 14.155038759689923,
"grad_norm": 0.34855517745018005,
"learning_rate": 0.0003879399488353305,
"loss": 3.0532,
"step": 52600
},
{
"epoch": 14.16849698535745,
"grad_norm": 0.3922046720981598,
"learning_rate": 0.0003877379830348727,
"loss": 3.0465,
"step": 52650
},
{
"epoch": 14.18195521102498,
"grad_norm": 0.3750133216381073,
"learning_rate": 0.0003875360172344149,
"loss": 3.0422,
"step": 52700
},
{
"epoch": 14.195413436692506,
"grad_norm": 0.3771599233150482,
"learning_rate": 0.0003873340514339571,
"loss": 3.0526,
"step": 52750
},
{
"epoch": 14.208871662360034,
"grad_norm": 0.3815842866897583,
"learning_rate": 0.0003871320856334993,
"loss": 3.0552,
"step": 52800
},
{
"epoch": 14.222329888027563,
"grad_norm": 0.4113651514053345,
"learning_rate": 0.00038693011983304156,
"loss": 3.057,
"step": 52850
},
{
"epoch": 14.23578811369509,
"grad_norm": 0.3831230401992798,
"learning_rate": 0.0003867281540325838,
"loss": 3.0523,
"step": 52900
},
{
"epoch": 14.24924633936262,
"grad_norm": 0.3881702423095703,
"learning_rate": 0.000386526188232126,
"loss": 3.0703,
"step": 52950
},
{
"epoch": 14.262704565030146,
"grad_norm": 0.3693937063217163,
"learning_rate": 0.00038632422243166825,
"loss": 3.0639,
"step": 53000
},
{
"epoch": 14.262704565030146,
"eval_accuracy": 0.3928689853643994,
"eval_loss": 3.3038170337677,
"eval_runtime": 53.2513,
"eval_samples_per_second": 338.246,
"eval_steps_per_second": 21.145,
"step": 53000
},
{
"epoch": 14.276162790697674,
"grad_norm": 0.3559059202671051,
"learning_rate": 0.00038612225663121045,
"loss": 3.0634,
"step": 53050
},
{
"epoch": 14.289621016365203,
"grad_norm": 0.4166322648525238,
"learning_rate": 0.00038592029083075265,
"loss": 3.0625,
"step": 53100
},
{
"epoch": 14.30307924203273,
"grad_norm": 0.3934720456600189,
"learning_rate": 0.00038571832503029484,
"loss": 3.0664,
"step": 53150
},
{
"epoch": 14.31653746770026,
"grad_norm": 0.38985350728034973,
"learning_rate": 0.00038551635922983704,
"loss": 3.0671,
"step": 53200
},
{
"epoch": 14.329995693367787,
"grad_norm": 0.3798811435699463,
"learning_rate": 0.0003853143934293793,
"loss": 3.0716,
"step": 53250
},
{
"epoch": 14.343453919035314,
"grad_norm": 0.3842821717262268,
"learning_rate": 0.0003851124276289215,
"loss": 3.066,
"step": 53300
},
{
"epoch": 14.356912144702843,
"grad_norm": 0.3599495589733124,
"learning_rate": 0.0003849104618284637,
"loss": 3.0733,
"step": 53350
},
{
"epoch": 14.37037037037037,
"grad_norm": 0.3783971965312958,
"learning_rate": 0.0003847084960280059,
"loss": 3.0649,
"step": 53400
},
{
"epoch": 14.383828596037898,
"grad_norm": 0.38178735971450806,
"learning_rate": 0.0003845065302275481,
"loss": 3.0699,
"step": 53450
},
{
"epoch": 14.397286821705427,
"grad_norm": 0.36713486909866333,
"learning_rate": 0.0003843045644270903,
"loss": 3.0643,
"step": 53500
},
{
"epoch": 14.410745047372954,
"grad_norm": 0.4043576717376709,
"learning_rate": 0.0003841025986266325,
"loss": 3.075,
"step": 53550
},
{
"epoch": 14.424203273040483,
"grad_norm": 0.3947139084339142,
"learning_rate": 0.0003839006328261747,
"loss": 3.077,
"step": 53600
},
{
"epoch": 14.43766149870801,
"grad_norm": 0.38625821471214294,
"learning_rate": 0.0003836986670257169,
"loss": 3.0728,
"step": 53650
},
{
"epoch": 14.451119724375538,
"grad_norm": 0.37357762455940247,
"learning_rate": 0.0003834967012252591,
"loss": 3.0738,
"step": 53700
},
{
"epoch": 14.464577950043067,
"grad_norm": 0.3948591649532318,
"learning_rate": 0.00038329473542480136,
"loss": 3.0722,
"step": 53750
},
{
"epoch": 14.478036175710594,
"grad_norm": 0.38419365882873535,
"learning_rate": 0.0003830927696243436,
"loss": 3.0827,
"step": 53800
},
{
"epoch": 14.491494401378123,
"grad_norm": 0.36976271867752075,
"learning_rate": 0.0003828908038238858,
"loss": 3.0789,
"step": 53850
},
{
"epoch": 14.50495262704565,
"grad_norm": 0.36416393518447876,
"learning_rate": 0.000382688838023428,
"loss": 3.0837,
"step": 53900
},
{
"epoch": 14.518410852713178,
"grad_norm": 0.3858191967010498,
"learning_rate": 0.00038248687222297025,
"loss": 3.0784,
"step": 53950
},
{
"epoch": 14.531869078380707,
"grad_norm": 0.404272198677063,
"learning_rate": 0.00038228490642251245,
"loss": 3.0816,
"step": 54000
},
{
"epoch": 14.531869078380707,
"eval_accuracy": 0.3930469489138139,
"eval_loss": 3.2975451946258545,
"eval_runtime": 53.2861,
"eval_samples_per_second": 338.024,
"eval_steps_per_second": 21.131,
"step": 54000
},
{
"epoch": 14.545327304048234,
"grad_norm": 0.37756025791168213,
"learning_rate": 0.00038208294062205464,
"loss": 3.0926,
"step": 54050
},
{
"epoch": 14.558785529715763,
"grad_norm": 0.40512412786483765,
"learning_rate": 0.00038188097482159684,
"loss": 3.0828,
"step": 54100
},
{
"epoch": 14.57224375538329,
"grad_norm": 0.3908289074897766,
"learning_rate": 0.00038167900902113904,
"loss": 3.0801,
"step": 54150
},
{
"epoch": 14.585701981050818,
"grad_norm": 0.4162972867488861,
"learning_rate": 0.0003814770432206813,
"loss": 3.0888,
"step": 54200
},
{
"epoch": 14.599160206718347,
"grad_norm": 0.4289568066596985,
"learning_rate": 0.0003812750774202235,
"loss": 3.0959,
"step": 54250
},
{
"epoch": 14.612618432385874,
"grad_norm": 0.3740384876728058,
"learning_rate": 0.0003810731116197657,
"loss": 3.0931,
"step": 54300
},
{
"epoch": 14.626076658053403,
"grad_norm": 0.36348482966423035,
"learning_rate": 0.0003808711458193079,
"loss": 3.0866,
"step": 54350
},
{
"epoch": 14.63953488372093,
"grad_norm": 0.39538612961769104,
"learning_rate": 0.00038066918001885007,
"loss": 3.0867,
"step": 54400
},
{
"epoch": 14.652993109388458,
"grad_norm": 0.3879004716873169,
"learning_rate": 0.0003804672142183923,
"loss": 3.0887,
"step": 54450
},
{
"epoch": 14.666451335055987,
"grad_norm": 0.355636864900589,
"learning_rate": 0.0003802652484179345,
"loss": 3.0893,
"step": 54500
},
{
"epoch": 14.679909560723514,
"grad_norm": 0.37391379475593567,
"learning_rate": 0.0003800632826174767,
"loss": 3.0987,
"step": 54550
},
{
"epoch": 14.693367786391041,
"grad_norm": 0.3761315643787384,
"learning_rate": 0.0003798613168170189,
"loss": 3.0868,
"step": 54600
},
{
"epoch": 14.70682601205857,
"grad_norm": 0.3604639172554016,
"learning_rate": 0.0003796593510165611,
"loss": 3.1005,
"step": 54650
},
{
"epoch": 14.720284237726098,
"grad_norm": 0.41159263253211975,
"learning_rate": 0.0003794573852161034,
"loss": 3.093,
"step": 54700
},
{
"epoch": 14.733742463393627,
"grad_norm": 0.39038246870040894,
"learning_rate": 0.0003792554194156456,
"loss": 3.0832,
"step": 54750
},
{
"epoch": 14.747200689061154,
"grad_norm": 0.39387059211730957,
"learning_rate": 0.0003790534536151878,
"loss": 3.1006,
"step": 54800
},
{
"epoch": 14.760658914728682,
"grad_norm": 0.37214162945747375,
"learning_rate": 0.00037885148781473005,
"loss": 3.094,
"step": 54850
},
{
"epoch": 14.77411714039621,
"grad_norm": 0.3764510154724121,
"learning_rate": 0.00037864952201427225,
"loss": 3.0977,
"step": 54900
},
{
"epoch": 14.787575366063738,
"grad_norm": 0.3526248037815094,
"learning_rate": 0.00037844755621381444,
"loss": 3.1015,
"step": 54950
},
{
"epoch": 14.801033591731267,
"grad_norm": 0.4617568552494049,
"learning_rate": 0.00037824559041335664,
"loss": 3.0951,
"step": 55000
},
{
"epoch": 14.801033591731267,
"eval_accuracy": 0.39400706117643686,
"eval_loss": 3.287712335586548,
"eval_runtime": 53.2104,
"eval_samples_per_second": 338.505,
"eval_steps_per_second": 21.161,
"step": 55000
},
{
"epoch": 14.814491817398794,
"grad_norm": 0.3996323347091675,
"learning_rate": 0.00037804362461289884,
"loss": 3.1008,
"step": 55050
},
{
"epoch": 14.827950043066322,
"grad_norm": 0.3723200261592865,
"learning_rate": 0.0003778416588124411,
"loss": 3.0927,
"step": 55100
},
{
"epoch": 14.84140826873385,
"grad_norm": 0.3873573839664459,
"learning_rate": 0.0003776396930119833,
"loss": 3.1062,
"step": 55150
},
{
"epoch": 14.854866494401378,
"grad_norm": 0.37784600257873535,
"learning_rate": 0.0003774377272115255,
"loss": 3.0936,
"step": 55200
},
{
"epoch": 14.868324720068905,
"grad_norm": 0.38209277391433716,
"learning_rate": 0.0003772357614110677,
"loss": 3.1048,
"step": 55250
},
{
"epoch": 14.881782945736434,
"grad_norm": 0.3873889446258545,
"learning_rate": 0.00037703379561060987,
"loss": 3.0883,
"step": 55300
},
{
"epoch": 14.895241171403962,
"grad_norm": 0.3716479241847992,
"learning_rate": 0.0003768318298101521,
"loss": 3.1102,
"step": 55350
},
{
"epoch": 14.90869939707149,
"grad_norm": 0.3743898570537567,
"learning_rate": 0.0003766298640096943,
"loss": 3.0971,
"step": 55400
},
{
"epoch": 14.922157622739018,
"grad_norm": 0.3971049189567566,
"learning_rate": 0.0003764278982092365,
"loss": 3.1034,
"step": 55450
},
{
"epoch": 14.935615848406545,
"grad_norm": 0.3767416179180145,
"learning_rate": 0.0003762259324087787,
"loss": 3.0957,
"step": 55500
},
{
"epoch": 14.949074074074074,
"grad_norm": 0.37302061915397644,
"learning_rate": 0.0003760239666083209,
"loss": 3.0999,
"step": 55550
},
{
"epoch": 14.962532299741602,
"grad_norm": 0.37083128094673157,
"learning_rate": 0.0003758220008078632,
"loss": 3.0928,
"step": 55600
},
{
"epoch": 14.97599052540913,
"grad_norm": 0.41234105825424194,
"learning_rate": 0.0003756200350074054,
"loss": 3.1026,
"step": 55650
},
{
"epoch": 14.989448751076658,
"grad_norm": 0.3714658319950104,
"learning_rate": 0.0003754180692069476,
"loss": 3.1092,
"step": 55700
},
{
"epoch": 15.002691645133506,
"grad_norm": 0.3906070590019226,
"learning_rate": 0.0003752161034064898,
"loss": 3.0814,
"step": 55750
},
{
"epoch": 15.016149870801033,
"grad_norm": 0.40009233355522156,
"learning_rate": 0.00037501413760603205,
"loss": 3.0027,
"step": 55800
},
{
"epoch": 15.02960809646856,
"grad_norm": 0.39637264609336853,
"learning_rate": 0.00037481217180557425,
"loss": 3.0125,
"step": 55850
},
{
"epoch": 15.04306632213609,
"grad_norm": 0.41015830636024475,
"learning_rate": 0.00037461020600511644,
"loss": 3.0059,
"step": 55900
},
{
"epoch": 15.056524547803617,
"grad_norm": 0.4093896746635437,
"learning_rate": 0.00037440824020465864,
"loss": 3.0224,
"step": 55950
},
{
"epoch": 15.069982773471146,
"grad_norm": 0.3793170750141144,
"learning_rate": 0.00037420627440420083,
"loss": 3.0158,
"step": 56000
},
{
"epoch": 15.069982773471146,
"eval_accuracy": 0.3932361030893516,
"eval_loss": 3.3005900382995605,
"eval_runtime": 53.2337,
"eval_samples_per_second": 338.357,
"eval_steps_per_second": 21.152,
"step": 56000
},
{
"epoch": 15.083440999138674,
"grad_norm": 0.3804849088191986,
"learning_rate": 0.0003740043086037431,
"loss": 3.0165,
"step": 56050
},
{
"epoch": 15.0968992248062,
"grad_norm": 0.41092732548713684,
"learning_rate": 0.0003738023428032853,
"loss": 3.0383,
"step": 56100
},
{
"epoch": 15.11035745047373,
"grad_norm": 0.39817407727241516,
"learning_rate": 0.0003736003770028275,
"loss": 3.0375,
"step": 56150
},
{
"epoch": 15.123815676141257,
"grad_norm": 0.4043419659137726,
"learning_rate": 0.00037339841120236967,
"loss": 3.0246,
"step": 56200
},
{
"epoch": 15.137273901808786,
"grad_norm": 0.39283016324043274,
"learning_rate": 0.00037319644540191187,
"loss": 3.0341,
"step": 56250
},
{
"epoch": 15.150732127476314,
"grad_norm": 0.3876219391822815,
"learning_rate": 0.0003729944796014541,
"loss": 3.0386,
"step": 56300
},
{
"epoch": 15.164190353143841,
"grad_norm": 0.39713042974472046,
"learning_rate": 0.0003727925138009963,
"loss": 3.0381,
"step": 56350
},
{
"epoch": 15.17764857881137,
"grad_norm": 0.3937109112739563,
"learning_rate": 0.0003725905480005385,
"loss": 3.0339,
"step": 56400
},
{
"epoch": 15.191106804478897,
"grad_norm": 0.41015708446502686,
"learning_rate": 0.0003723885822000807,
"loss": 3.0223,
"step": 56450
},
{
"epoch": 15.204565030146426,
"grad_norm": 0.4246147572994232,
"learning_rate": 0.000372186616399623,
"loss": 3.0406,
"step": 56500
},
{
"epoch": 15.218023255813954,
"grad_norm": 0.4137839376926422,
"learning_rate": 0.0003719846505991652,
"loss": 3.038,
"step": 56550
},
{
"epoch": 15.231481481481481,
"grad_norm": 0.3703172206878662,
"learning_rate": 0.0003717826847987074,
"loss": 3.043,
"step": 56600
},
{
"epoch": 15.24493970714901,
"grad_norm": 0.3823303282260895,
"learning_rate": 0.0003715807189982496,
"loss": 3.0399,
"step": 56650
},
{
"epoch": 15.258397932816537,
"grad_norm": 0.4032827317714691,
"learning_rate": 0.00037137875319779185,
"loss": 3.0372,
"step": 56700
},
{
"epoch": 15.271856158484065,
"grad_norm": 0.3879244029521942,
"learning_rate": 0.00037117678739733405,
"loss": 3.0386,
"step": 56750
},
{
"epoch": 15.285314384151594,
"grad_norm": 0.42025917768478394,
"learning_rate": 0.00037097482159687624,
"loss": 3.0495,
"step": 56800
},
{
"epoch": 15.298772609819121,
"grad_norm": 0.4079301357269287,
"learning_rate": 0.00037077285579641844,
"loss": 3.0522,
"step": 56850
},
{
"epoch": 15.31223083548665,
"grad_norm": 0.3735152781009674,
"learning_rate": 0.00037057088999596063,
"loss": 3.0518,
"step": 56900
},
{
"epoch": 15.325689061154177,
"grad_norm": 0.3667006492614746,
"learning_rate": 0.0003703689241955029,
"loss": 3.0653,
"step": 56950
},
{
"epoch": 15.339147286821705,
"grad_norm": 0.3790123760700226,
"learning_rate": 0.0003701669583950451,
"loss": 3.055,
"step": 57000
},
{
"epoch": 15.339147286821705,
"eval_accuracy": 0.3934124369359327,
"eval_loss": 3.2985215187072754,
"eval_runtime": 53.2792,
"eval_samples_per_second": 338.068,
"eval_steps_per_second": 21.134,
"step": 57000
},
{
"epoch": 15.352605512489234,
"grad_norm": 0.39140379428863525,
"learning_rate": 0.0003699649925945873,
"loss": 3.0502,
"step": 57050
},
{
"epoch": 15.366063738156761,
"grad_norm": 0.36787888407707214,
"learning_rate": 0.0003697630267941295,
"loss": 3.0509,
"step": 57100
},
{
"epoch": 15.37952196382429,
"grad_norm": 0.398590624332428,
"learning_rate": 0.00036956106099367167,
"loss": 3.0607,
"step": 57150
},
{
"epoch": 15.392980189491817,
"grad_norm": 0.40813106298446655,
"learning_rate": 0.0003693590951932139,
"loss": 3.0614,
"step": 57200
},
{
"epoch": 15.406438415159345,
"grad_norm": 0.4021749496459961,
"learning_rate": 0.0003691571293927561,
"loss": 3.072,
"step": 57250
},
{
"epoch": 15.419896640826874,
"grad_norm": 0.4174582064151764,
"learning_rate": 0.0003689551635922983,
"loss": 3.0584,
"step": 57300
},
{
"epoch": 15.433354866494401,
"grad_norm": 0.3853360712528229,
"learning_rate": 0.00036875319779184056,
"loss": 3.0664,
"step": 57350
},
{
"epoch": 15.44681309216193,
"grad_norm": 0.39959678053855896,
"learning_rate": 0.0003685512319913828,
"loss": 3.0574,
"step": 57400
},
{
"epoch": 15.460271317829458,
"grad_norm": 0.3768465220928192,
"learning_rate": 0.000368349266190925,
"loss": 3.064,
"step": 57450
},
{
"epoch": 15.473729543496985,
"grad_norm": 0.413850873708725,
"learning_rate": 0.0003681473003904672,
"loss": 3.058,
"step": 57500
},
{
"epoch": 15.487187769164514,
"grad_norm": 0.3981676697731018,
"learning_rate": 0.0003679453345900094,
"loss": 3.0743,
"step": 57550
},
{
"epoch": 15.500645994832041,
"grad_norm": 0.3920087516307831,
"learning_rate": 0.0003677433687895516,
"loss": 3.0711,
"step": 57600
},
{
"epoch": 15.514104220499568,
"grad_norm": 0.390938937664032,
"learning_rate": 0.00036754140298909385,
"loss": 3.0714,
"step": 57650
},
{
"epoch": 15.527562446167098,
"grad_norm": 0.37027981877326965,
"learning_rate": 0.00036733943718863604,
"loss": 3.0682,
"step": 57700
},
{
"epoch": 15.541020671834625,
"grad_norm": 0.3791489899158478,
"learning_rate": 0.00036713747138817824,
"loss": 3.0739,
"step": 57750
},
{
"epoch": 15.554478897502154,
"grad_norm": 0.4081215560436249,
"learning_rate": 0.00036693550558772044,
"loss": 3.0656,
"step": 57800
},
{
"epoch": 15.567937123169681,
"grad_norm": 0.41583451628685,
"learning_rate": 0.00036673353978726263,
"loss": 3.0664,
"step": 57850
},
{
"epoch": 15.581395348837209,
"grad_norm": 0.37946993112564087,
"learning_rate": 0.0003665315739868049,
"loss": 3.0627,
"step": 57900
},
{
"epoch": 15.594853574504738,
"grad_norm": 0.37393489480018616,
"learning_rate": 0.0003663296081863471,
"loss": 3.067,
"step": 57950
},
{
"epoch": 15.608311800172265,
"grad_norm": 0.3997141122817993,
"learning_rate": 0.0003661276423858893,
"loss": 3.0719,
"step": 58000
},
{
"epoch": 15.608311800172265,
"eval_accuracy": 0.3940487815689736,
"eval_loss": 3.291177749633789,
"eval_runtime": 53.087,
"eval_samples_per_second": 339.292,
"eval_steps_per_second": 21.21,
"step": 58000
},
{
"epoch": 15.621770025839794,
"grad_norm": 0.3671659827232361,
"learning_rate": 0.00036592567658543147,
"loss": 3.0725,
"step": 58050
},
{
"epoch": 15.635228251507321,
"grad_norm": 0.3693672716617584,
"learning_rate": 0.00036572371078497367,
"loss": 3.0773,
"step": 58100
},
{
"epoch": 15.648686477174849,
"grad_norm": 0.37850597500801086,
"learning_rate": 0.0003655217449845159,
"loss": 3.0868,
"step": 58150
},
{
"epoch": 15.662144702842378,
"grad_norm": 0.3677229583263397,
"learning_rate": 0.0003653197791840581,
"loss": 3.0793,
"step": 58200
},
{
"epoch": 15.675602928509905,
"grad_norm": 0.3778005838394165,
"learning_rate": 0.00036511781338360036,
"loss": 3.0812,
"step": 58250
},
{
"epoch": 15.689061154177434,
"grad_norm": 0.38959360122680664,
"learning_rate": 0.00036491584758314256,
"loss": 3.0832,
"step": 58300
},
{
"epoch": 15.702519379844961,
"grad_norm": 0.3569706082344055,
"learning_rate": 0.0003647138817826848,
"loss": 3.0764,
"step": 58350
},
{
"epoch": 15.715977605512489,
"grad_norm": 0.392900675535202,
"learning_rate": 0.000364511915982227,
"loss": 3.0782,
"step": 58400
},
{
"epoch": 15.729435831180018,
"grad_norm": 0.3854053318500519,
"learning_rate": 0.0003643099501817692,
"loss": 3.0711,
"step": 58450
},
{
"epoch": 15.742894056847545,
"grad_norm": 0.42321139574050903,
"learning_rate": 0.0003641079843813114,
"loss": 3.0864,
"step": 58500
},
{
"epoch": 15.756352282515074,
"grad_norm": 0.3787823021411896,
"learning_rate": 0.0003639060185808536,
"loss": 3.0788,
"step": 58550
},
{
"epoch": 15.769810508182601,
"grad_norm": 0.4046003818511963,
"learning_rate": 0.00036370405278039584,
"loss": 3.0783,
"step": 58600
},
{
"epoch": 15.783268733850129,
"grad_norm": 0.3918498754501343,
"learning_rate": 0.00036350208697993804,
"loss": 3.0908,
"step": 58650
},
{
"epoch": 15.796726959517658,
"grad_norm": 0.38387343287467957,
"learning_rate": 0.00036330012117948024,
"loss": 3.0864,
"step": 58700
},
{
"epoch": 15.810185185185185,
"grad_norm": 0.35123082995414734,
"learning_rate": 0.00036309815537902243,
"loss": 3.0814,
"step": 58750
},
{
"epoch": 15.823643410852712,
"grad_norm": 0.3968450129032135,
"learning_rate": 0.0003628961895785647,
"loss": 3.0692,
"step": 58800
},
{
"epoch": 15.837101636520241,
"grad_norm": 0.3727889358997345,
"learning_rate": 0.0003626942237781069,
"loss": 3.0876,
"step": 58850
},
{
"epoch": 15.850559862187769,
"grad_norm": 0.363513708114624,
"learning_rate": 0.0003624922579776491,
"loss": 3.0858,
"step": 58900
},
{
"epoch": 15.864018087855298,
"grad_norm": 0.4061371386051178,
"learning_rate": 0.00036229029217719127,
"loss": 3.0875,
"step": 58950
},
{
"epoch": 15.877476313522825,
"grad_norm": 0.39571666717529297,
"learning_rate": 0.00036208832637673347,
"loss": 3.0866,
"step": 59000
},
{
"epoch": 15.877476313522825,
"eval_accuracy": 0.3941058211681449,
"eval_loss": 3.2855560779571533,
"eval_runtime": 53.4013,
"eval_samples_per_second": 337.295,
"eval_steps_per_second": 21.086,
"step": 59000
},
{
"epoch": 15.890934539190352,
"grad_norm": 0.3872230350971222,
"learning_rate": 0.0003618863605762757,
"loss": 3.085,
"step": 59050
},
{
"epoch": 15.904392764857882,
"grad_norm": 0.3803175687789917,
"learning_rate": 0.0003616843947758179,
"loss": 3.0744,
"step": 59100
},
{
"epoch": 15.917850990525409,
"grad_norm": 0.3779771029949188,
"learning_rate": 0.00036148242897536016,
"loss": 3.0824,
"step": 59150
},
{
"epoch": 15.931309216192938,
"grad_norm": 0.37976348400115967,
"learning_rate": 0.00036128046317490236,
"loss": 3.0853,
"step": 59200
},
{
"epoch": 15.944767441860465,
"grad_norm": 0.37095963954925537,
"learning_rate": 0.0003610784973744446,
"loss": 3.0841,
"step": 59250
},
{
"epoch": 15.958225667527993,
"grad_norm": 0.3925478458404541,
"learning_rate": 0.0003608765315739868,
"loss": 3.0902,
"step": 59300
},
{
"epoch": 15.971683893195522,
"grad_norm": 0.3674505650997162,
"learning_rate": 0.000360674565773529,
"loss": 3.0851,
"step": 59350
},
{
"epoch": 15.985142118863049,
"grad_norm": 0.3850049674510956,
"learning_rate": 0.0003604725999730712,
"loss": 3.0937,
"step": 59400
},
{
"epoch": 15.998600344530576,
"grad_norm": 0.3854713439941406,
"learning_rate": 0.0003602706341726134,
"loss": 3.0876,
"step": 59450
},
{
"epoch": 16.011843238587424,
"grad_norm": 0.39652571082115173,
"learning_rate": 0.00036006866837215564,
"loss": 3.0026,
"step": 59500
},
{
"epoch": 16.02530146425495,
"grad_norm": 0.39200231432914734,
"learning_rate": 0.00035986670257169784,
"loss": 2.9887,
"step": 59550
},
{
"epoch": 16.03875968992248,
"grad_norm": 0.3969775438308716,
"learning_rate": 0.00035966473677124004,
"loss": 2.9961,
"step": 59600
},
{
"epoch": 16.05221791559001,
"grad_norm": 0.3958861231803894,
"learning_rate": 0.00035946277097078223,
"loss": 3.0127,
"step": 59650
},
{
"epoch": 16.065676141257537,
"grad_norm": 0.4136175215244293,
"learning_rate": 0.00035926080517032443,
"loss": 2.995,
"step": 59700
},
{
"epoch": 16.079134366925064,
"grad_norm": 0.39139825105667114,
"learning_rate": 0.0003590588393698667,
"loss": 3.0106,
"step": 59750
},
{
"epoch": 16.09259259259259,
"grad_norm": 0.3823096454143524,
"learning_rate": 0.0003588568735694089,
"loss": 3.0173,
"step": 59800
},
{
"epoch": 16.10605081826012,
"grad_norm": 0.4026496112346649,
"learning_rate": 0.00035865490776895107,
"loss": 3.0072,
"step": 59850
},
{
"epoch": 16.11950904392765,
"grad_norm": 0.37121763825416565,
"learning_rate": 0.00035845294196849327,
"loss": 3.0127,
"step": 59900
},
{
"epoch": 16.132967269595177,
"grad_norm": 0.3773496150970459,
"learning_rate": 0.00035825097616803546,
"loss": 3.0064,
"step": 59950
},
{
"epoch": 16.146425495262704,
"grad_norm": 0.40183204412460327,
"learning_rate": 0.0003580490103675777,
"loss": 3.0163,
"step": 60000
},
{
"epoch": 16.146425495262704,
"eval_accuracy": 0.3936537416021413,
"eval_loss": 3.2996578216552734,
"eval_runtime": 53.3516,
"eval_samples_per_second": 337.609,
"eval_steps_per_second": 21.105,
"step": 60000
},
{
"epoch": 16.15988372093023,
"grad_norm": 0.41072332859039307,
"learning_rate": 0.00035784704456711996,
"loss": 3.0291,
"step": 60050
},
{
"epoch": 16.17334194659776,
"grad_norm": 0.3828774094581604,
"learning_rate": 0.00035764507876666216,
"loss": 3.0217,
"step": 60100
},
{
"epoch": 16.18680017226529,
"grad_norm": 0.37735939025878906,
"learning_rate": 0.00035744311296620436,
"loss": 3.0309,
"step": 60150
},
{
"epoch": 16.200258397932817,
"grad_norm": 0.41258010268211365,
"learning_rate": 0.0003572411471657466,
"loss": 3.022,
"step": 60200
},
{
"epoch": 16.213716623600344,
"grad_norm": 0.426175057888031,
"learning_rate": 0.0003570391813652888,
"loss": 3.0247,
"step": 60250
},
{
"epoch": 16.227174849267872,
"grad_norm": 0.3841751217842102,
"learning_rate": 0.000356837215564831,
"loss": 3.0353,
"step": 60300
},
{
"epoch": 16.2406330749354,
"grad_norm": 0.3686473071575165,
"learning_rate": 0.0003566352497643732,
"loss": 3.0349,
"step": 60350
},
{
"epoch": 16.25409130060293,
"grad_norm": 0.39995282888412476,
"learning_rate": 0.0003564332839639154,
"loss": 3.0311,
"step": 60400
},
{
"epoch": 16.267549526270457,
"grad_norm": 0.3944709002971649,
"learning_rate": 0.00035623131816345764,
"loss": 3.0363,
"step": 60450
},
{
"epoch": 16.281007751937985,
"grad_norm": 0.42645901441574097,
"learning_rate": 0.00035602935236299984,
"loss": 3.0346,
"step": 60500
},
{
"epoch": 16.294465977605512,
"grad_norm": 0.3777204751968384,
"learning_rate": 0.00035582738656254203,
"loss": 3.0379,
"step": 60550
},
{
"epoch": 16.30792420327304,
"grad_norm": 0.43480342626571655,
"learning_rate": 0.00035562542076208423,
"loss": 3.0456,
"step": 60600
},
{
"epoch": 16.32138242894057,
"grad_norm": 0.41234782338142395,
"learning_rate": 0.0003554234549616264,
"loss": 3.0441,
"step": 60650
},
{
"epoch": 16.334840654608097,
"grad_norm": 0.4231054484844208,
"learning_rate": 0.0003552214891611687,
"loss": 3.0439,
"step": 60700
},
{
"epoch": 16.348298880275625,
"grad_norm": 0.40122222900390625,
"learning_rate": 0.00035501952336071087,
"loss": 3.043,
"step": 60750
},
{
"epoch": 16.361757105943152,
"grad_norm": 0.39891695976257324,
"learning_rate": 0.00035481755756025307,
"loss": 3.0397,
"step": 60800
},
{
"epoch": 16.37521533161068,
"grad_norm": 0.40936100482940674,
"learning_rate": 0.00035461559175979526,
"loss": 3.0488,
"step": 60850
},
{
"epoch": 16.38867355727821,
"grad_norm": 0.4073893129825592,
"learning_rate": 0.0003544136259593375,
"loss": 3.0487,
"step": 60900
},
{
"epoch": 16.402131782945737,
"grad_norm": 0.3983805477619171,
"learning_rate": 0.00035421166015887977,
"loss": 3.0557,
"step": 60950
},
{
"epoch": 16.415590008613265,
"grad_norm": 0.39748522639274597,
"learning_rate": 0.00035400969435842196,
"loss": 3.0327,
"step": 61000
},
{
"epoch": 16.415590008613265,
"eval_accuracy": 0.3939599084411219,
"eval_loss": 3.2969777584075928,
"eval_runtime": 53.184,
"eval_samples_per_second": 338.673,
"eval_steps_per_second": 21.172,
"step": 61000
},
{
"epoch": 16.429048234280792,
"grad_norm": 0.4088171720504761,
"learning_rate": 0.00035380772855796416,
"loss": 3.0601,
"step": 61050
},
{
"epoch": 16.44250645994832,
"grad_norm": 0.3807612359523773,
"learning_rate": 0.0003536057627575064,
"loss": 3.0543,
"step": 61100
},
{
"epoch": 16.45596468561585,
"grad_norm": 0.4410359859466553,
"learning_rate": 0.0003534037969570486,
"loss": 3.0499,
"step": 61150
},
{
"epoch": 16.469422911283377,
"grad_norm": 0.40752777457237244,
"learning_rate": 0.0003532018311565908,
"loss": 3.0546,
"step": 61200
},
{
"epoch": 16.482881136950905,
"grad_norm": 0.38025355339050293,
"learning_rate": 0.000352999865356133,
"loss": 3.0511,
"step": 61250
},
{
"epoch": 16.496339362618432,
"grad_norm": 0.38205963373184204,
"learning_rate": 0.0003527978995556752,
"loss": 3.0504,
"step": 61300
},
{
"epoch": 16.50979758828596,
"grad_norm": 0.41273659467697144,
"learning_rate": 0.00035259593375521744,
"loss": 3.0582,
"step": 61350
},
{
"epoch": 16.52325581395349,
"grad_norm": 0.43403419852256775,
"learning_rate": 0.00035239396795475964,
"loss": 3.0558,
"step": 61400
},
{
"epoch": 16.536714039621017,
"grad_norm": 0.4052940309047699,
"learning_rate": 0.00035219200215430183,
"loss": 3.064,
"step": 61450
},
{
"epoch": 16.550172265288545,
"grad_norm": 0.4491784870624542,
"learning_rate": 0.00035199003635384403,
"loss": 3.0662,
"step": 61500
},
{
"epoch": 16.563630490956072,
"grad_norm": 0.40795519948005676,
"learning_rate": 0.0003517880705533862,
"loss": 3.0541,
"step": 61550
},
{
"epoch": 16.5770887166236,
"grad_norm": 0.4100624620914459,
"learning_rate": 0.0003515861047529285,
"loss": 3.0603,
"step": 61600
},
{
"epoch": 16.590546942291127,
"grad_norm": 0.3613821864128113,
"learning_rate": 0.0003513841389524707,
"loss": 3.052,
"step": 61650
},
{
"epoch": 16.604005167958658,
"grad_norm": 0.399851530790329,
"learning_rate": 0.00035118217315201287,
"loss": 3.0598,
"step": 61700
},
{
"epoch": 16.617463393626185,
"grad_norm": 0.37738752365112305,
"learning_rate": 0.00035098020735155507,
"loss": 3.0629,
"step": 61750
},
{
"epoch": 16.630921619293712,
"grad_norm": 0.4187788665294647,
"learning_rate": 0.00035077824155109737,
"loss": 3.0551,
"step": 61800
},
{
"epoch": 16.64437984496124,
"grad_norm": 0.389838308095932,
"learning_rate": 0.00035057627575063957,
"loss": 3.0685,
"step": 61850
},
{
"epoch": 16.657838070628767,
"grad_norm": 0.41789937019348145,
"learning_rate": 0.00035037430995018176,
"loss": 3.0572,
"step": 61900
},
{
"epoch": 16.671296296296298,
"grad_norm": 0.38143324851989746,
"learning_rate": 0.00035017234414972396,
"loss": 3.0631,
"step": 61950
},
{
"epoch": 16.684754521963825,
"grad_norm": 0.3955875039100647,
"learning_rate": 0.00034997037834926615,
"loss": 3.0601,
"step": 62000
},
{
"epoch": 16.684754521963825,
"eval_accuracy": 0.39469718600298215,
"eval_loss": 3.2880349159240723,
"eval_runtime": 53.1791,
"eval_samples_per_second": 338.704,
"eval_steps_per_second": 21.174,
"step": 62000
},
{
"epoch": 16.698212747631352,
"grad_norm": 0.3697275221347809,
"learning_rate": 0.0003497684125488084,
"loss": 3.0608,
"step": 62050
},
{
"epoch": 16.71167097329888,
"grad_norm": 0.39858055114746094,
"learning_rate": 0.0003495664467483506,
"loss": 3.0612,
"step": 62100
},
{
"epoch": 16.725129198966407,
"grad_norm": 0.4065309762954712,
"learning_rate": 0.0003493644809478928,
"loss": 3.0739,
"step": 62150
},
{
"epoch": 16.738587424633938,
"grad_norm": 0.39412033557891846,
"learning_rate": 0.000349162515147435,
"loss": 3.063,
"step": 62200
},
{
"epoch": 16.752045650301465,
"grad_norm": 0.37209275364875793,
"learning_rate": 0.0003489605493469772,
"loss": 3.0661,
"step": 62250
},
{
"epoch": 16.765503875968992,
"grad_norm": 0.40857645869255066,
"learning_rate": 0.00034875858354651944,
"loss": 3.057,
"step": 62300
},
{
"epoch": 16.77896210163652,
"grad_norm": 0.3995668888092041,
"learning_rate": 0.00034855661774606164,
"loss": 3.0631,
"step": 62350
},
{
"epoch": 16.792420327304047,
"grad_norm": 0.4129258990287781,
"learning_rate": 0.00034835465194560383,
"loss": 3.0786,
"step": 62400
},
{
"epoch": 16.805878552971578,
"grad_norm": 0.37398791313171387,
"learning_rate": 0.00034815268614514603,
"loss": 3.0683,
"step": 62450
},
{
"epoch": 16.819336778639105,
"grad_norm": 0.3871100842952728,
"learning_rate": 0.0003479507203446882,
"loss": 3.0685,
"step": 62500
},
{
"epoch": 16.832795004306632,
"grad_norm": 0.3890288472175598,
"learning_rate": 0.0003477487545442305,
"loss": 3.0733,
"step": 62550
},
{
"epoch": 16.84625322997416,
"grad_norm": 0.4500114321708679,
"learning_rate": 0.00034754678874377267,
"loss": 3.0633,
"step": 62600
},
{
"epoch": 16.859711455641687,
"grad_norm": 0.3752024173736572,
"learning_rate": 0.00034734482294331487,
"loss": 3.0744,
"step": 62650
},
{
"epoch": 16.873169681309218,
"grad_norm": 0.3815058469772339,
"learning_rate": 0.00034714285714285717,
"loss": 3.0623,
"step": 62700
},
{
"epoch": 16.886627906976745,
"grad_norm": 0.41180121898651123,
"learning_rate": 0.00034694089134239937,
"loss": 3.0753,
"step": 62750
},
{
"epoch": 16.900086132644272,
"grad_norm": 0.40455207228660583,
"learning_rate": 0.00034673892554194156,
"loss": 3.0696,
"step": 62800
},
{
"epoch": 16.9135443583118,
"grad_norm": 0.40730997920036316,
"learning_rate": 0.00034653695974148376,
"loss": 3.0663,
"step": 62850
},
{
"epoch": 16.927002583979327,
"grad_norm": 0.4231722056865692,
"learning_rate": 0.00034633499394102595,
"loss": 3.0777,
"step": 62900
},
{
"epoch": 16.940460809646858,
"grad_norm": 0.39433252811431885,
"learning_rate": 0.0003461330281405682,
"loss": 3.0782,
"step": 62950
},
{
"epoch": 16.953919035314385,
"grad_norm": 0.38897091150283813,
"learning_rate": 0.0003459310623401104,
"loss": 3.0832,
"step": 63000
},
{
"epoch": 16.953919035314385,
"eval_accuracy": 0.3949562001066477,
"eval_loss": 3.282935857772827,
"eval_runtime": 53.1939,
"eval_samples_per_second": 338.61,
"eval_steps_per_second": 21.168,
"step": 63000
},
{
"epoch": 16.967377260981912,
"grad_norm": 0.3950527608394623,
"learning_rate": 0.0003457290965396526,
"loss": 3.0736,
"step": 63050
},
{
"epoch": 16.98083548664944,
"grad_norm": 0.38114306330680847,
"learning_rate": 0.0003455271307391948,
"loss": 3.0815,
"step": 63100
},
{
"epoch": 16.994293712316967,
"grad_norm": 0.3700971305370331,
"learning_rate": 0.000345325164938737,
"loss": 3.0773,
"step": 63150
},
{
"epoch": 17.007536606373815,
"grad_norm": 0.4128997325897217,
"learning_rate": 0.00034512319913827924,
"loss": 3.0138,
"step": 63200
},
{
"epoch": 17.020994832041342,
"grad_norm": 0.3955026865005493,
"learning_rate": 0.00034492123333782144,
"loss": 2.9982,
"step": 63250
},
{
"epoch": 17.034453057708873,
"grad_norm": 0.38770145177841187,
"learning_rate": 0.00034471926753736363,
"loss": 2.9895,
"step": 63300
},
{
"epoch": 17.0479112833764,
"grad_norm": 0.46743109822273254,
"learning_rate": 0.00034451730173690583,
"loss": 2.9831,
"step": 63350
},
{
"epoch": 17.061369509043928,
"grad_norm": 0.3955764174461365,
"learning_rate": 0.000344315335936448,
"loss": 2.9904,
"step": 63400
},
{
"epoch": 17.074827734711455,
"grad_norm": 0.36407607793807983,
"learning_rate": 0.0003441133701359903,
"loss": 2.9922,
"step": 63450
},
{
"epoch": 17.088285960378983,
"grad_norm": 0.40364158153533936,
"learning_rate": 0.00034391140433553247,
"loss": 2.9953,
"step": 63500
},
{
"epoch": 17.101744186046513,
"grad_norm": 0.4021804630756378,
"learning_rate": 0.00034370943853507467,
"loss": 2.9977,
"step": 63550
},
{
"epoch": 17.11520241171404,
"grad_norm": 0.37996798753738403,
"learning_rate": 0.0003435074727346169,
"loss": 3.0054,
"step": 63600
},
{
"epoch": 17.128660637381568,
"grad_norm": 0.4075641930103302,
"learning_rate": 0.00034330550693415917,
"loss": 3.0035,
"step": 63650
},
{
"epoch": 17.142118863049095,
"grad_norm": 0.40062153339385986,
"learning_rate": 0.00034310354113370136,
"loss": 2.9982,
"step": 63700
},
{
"epoch": 17.155577088716623,
"grad_norm": 0.4038476049900055,
"learning_rate": 0.00034290157533324356,
"loss": 3.0037,
"step": 63750
},
{
"epoch": 17.16903531438415,
"grad_norm": 0.3905138373374939,
"learning_rate": 0.00034269960953278576,
"loss": 3.0206,
"step": 63800
},
{
"epoch": 17.18249354005168,
"grad_norm": 0.4093771278858185,
"learning_rate": 0.00034249764373232795,
"loss": 3.0089,
"step": 63850
},
{
"epoch": 17.195951765719208,
"grad_norm": 0.3983815014362335,
"learning_rate": 0.0003422956779318702,
"loss": 3.0094,
"step": 63900
},
{
"epoch": 17.209409991386735,
"grad_norm": 0.39827215671539307,
"learning_rate": 0.0003420937121314124,
"loss": 3.0127,
"step": 63950
},
{
"epoch": 17.222868217054263,
"grad_norm": 0.4104834794998169,
"learning_rate": 0.0003418917463309546,
"loss": 3.0158,
"step": 64000
},
{
"epoch": 17.222868217054263,
"eval_accuracy": 0.39382040587857714,
"eval_loss": 3.2985177040100098,
"eval_runtime": 53.135,
"eval_samples_per_second": 338.986,
"eval_steps_per_second": 21.191,
"step": 64000
},
{
"epoch": 17.23632644272179,
"grad_norm": 0.38256514072418213,
"learning_rate": 0.0003416897805304968,
"loss": 3.0228,
"step": 64050
},
{
"epoch": 17.24978466838932,
"grad_norm": 0.38628092408180237,
"learning_rate": 0.000341487814730039,
"loss": 3.0113,
"step": 64100
},
{
"epoch": 17.263242894056848,
"grad_norm": 0.3880211114883423,
"learning_rate": 0.00034128584892958124,
"loss": 3.0162,
"step": 64150
},
{
"epoch": 17.276701119724375,
"grad_norm": 0.3892453610897064,
"learning_rate": 0.00034108388312912343,
"loss": 3.0153,
"step": 64200
},
{
"epoch": 17.290159345391903,
"grad_norm": 0.3993564248085022,
"learning_rate": 0.00034088191732866563,
"loss": 3.0313,
"step": 64250
},
{
"epoch": 17.30361757105943,
"grad_norm": 0.3808676600456238,
"learning_rate": 0.0003406799515282078,
"loss": 3.0332,
"step": 64300
},
{
"epoch": 17.31707579672696,
"grad_norm": 0.39774927496910095,
"learning_rate": 0.00034047798572775,
"loss": 3.0197,
"step": 64350
},
{
"epoch": 17.330534022394488,
"grad_norm": 0.4316226840019226,
"learning_rate": 0.00034027601992729227,
"loss": 3.0257,
"step": 64400
},
{
"epoch": 17.343992248062015,
"grad_norm": 0.42357945442199707,
"learning_rate": 0.00034007405412683447,
"loss": 3.041,
"step": 64450
},
{
"epoch": 17.357450473729543,
"grad_norm": 0.4115462899208069,
"learning_rate": 0.0003398720883263767,
"loss": 3.0264,
"step": 64500
},
{
"epoch": 17.37090869939707,
"grad_norm": 0.3940788209438324,
"learning_rate": 0.00033967012252591897,
"loss": 3.0349,
"step": 64550
},
{
"epoch": 17.3843669250646,
"grad_norm": 0.48263993859291077,
"learning_rate": 0.00033946815672546116,
"loss": 3.0464,
"step": 64600
},
{
"epoch": 17.397825150732128,
"grad_norm": 0.41211599111557007,
"learning_rate": 0.00033926619092500336,
"loss": 3.0267,
"step": 64650
},
{
"epoch": 17.411283376399656,
"grad_norm": 0.4162316918373108,
"learning_rate": 0.00033906422512454556,
"loss": 3.0242,
"step": 64700
},
{
"epoch": 17.424741602067183,
"grad_norm": 0.4139719009399414,
"learning_rate": 0.00033886225932408775,
"loss": 3.038,
"step": 64750
},
{
"epoch": 17.43819982773471,
"grad_norm": 0.38862109184265137,
"learning_rate": 0.00033866029352363,
"loss": 3.0433,
"step": 64800
},
{
"epoch": 17.45165805340224,
"grad_norm": 0.4061868190765381,
"learning_rate": 0.0003384583277231722,
"loss": 3.0346,
"step": 64850
},
{
"epoch": 17.46511627906977,
"grad_norm": 0.3974857032299042,
"learning_rate": 0.0003382563619227144,
"loss": 3.0427,
"step": 64900
},
{
"epoch": 17.478574504737296,
"grad_norm": 0.37685614824295044,
"learning_rate": 0.0003380543961222566,
"loss": 3.0387,
"step": 64950
},
{
"epoch": 17.492032730404823,
"grad_norm": 0.4107515215873718,
"learning_rate": 0.0003378524303217988,
"loss": 3.0447,
"step": 65000
},
{
"epoch": 17.492032730404823,
"eval_accuracy": 0.3945491003388478,
"eval_loss": 3.2916104793548584,
"eval_runtime": 53.1618,
"eval_samples_per_second": 338.814,
"eval_steps_per_second": 21.181,
"step": 65000
},
{
"epoch": 17.50549095607235,
"grad_norm": 0.4122573733329773,
"learning_rate": 0.00033765046452134104,
"loss": 3.04,
"step": 65050
},
{
"epoch": 17.51894918173988,
"grad_norm": 0.37764185667037964,
"learning_rate": 0.00033744849872088323,
"loss": 3.04,
"step": 65100
},
{
"epoch": 17.53240740740741,
"grad_norm": 0.39302051067352295,
"learning_rate": 0.00033724653292042543,
"loss": 3.0438,
"step": 65150
},
{
"epoch": 17.545865633074936,
"grad_norm": 0.41048935055732727,
"learning_rate": 0.0003370445671199676,
"loss": 3.0469,
"step": 65200
},
{
"epoch": 17.559323858742463,
"grad_norm": 0.4067849814891815,
"learning_rate": 0.0003368426013195098,
"loss": 3.0454,
"step": 65250
},
{
"epoch": 17.57278208440999,
"grad_norm": 0.42745301127433777,
"learning_rate": 0.00033664063551905207,
"loss": 3.0354,
"step": 65300
},
{
"epoch": 17.58624031007752,
"grad_norm": 0.40041837096214294,
"learning_rate": 0.00033643866971859427,
"loss": 3.0553,
"step": 65350
},
{
"epoch": 17.59969853574505,
"grad_norm": 0.40067821741104126,
"learning_rate": 0.0003362367039181365,
"loss": 3.0344,
"step": 65400
},
{
"epoch": 17.613156761412576,
"grad_norm": 0.41964805126190186,
"learning_rate": 0.0003360347381176787,
"loss": 3.0544,
"step": 65450
},
{
"epoch": 17.626614987080103,
"grad_norm": 0.4025363028049469,
"learning_rate": 0.00033583277231722097,
"loss": 3.0454,
"step": 65500
},
{
"epoch": 17.64007321274763,
"grad_norm": 0.4273861050605774,
"learning_rate": 0.00033563080651676316,
"loss": 3.0504,
"step": 65550
},
{
"epoch": 17.653531438415158,
"grad_norm": 0.4114350378513336,
"learning_rate": 0.00033542884071630536,
"loss": 3.0515,
"step": 65600
},
{
"epoch": 17.66698966408269,
"grad_norm": 0.38349467515945435,
"learning_rate": 0.00033522687491584755,
"loss": 3.0493,
"step": 65650
},
{
"epoch": 17.680447889750216,
"grad_norm": 0.38878414034843445,
"learning_rate": 0.00033502490911538975,
"loss": 3.0462,
"step": 65700
},
{
"epoch": 17.693906115417743,
"grad_norm": 0.39213553071022034,
"learning_rate": 0.000334822943314932,
"loss": 3.0496,
"step": 65750
},
{
"epoch": 17.70736434108527,
"grad_norm": 0.40689966082572937,
"learning_rate": 0.0003346209775144742,
"loss": 3.0513,
"step": 65800
},
{
"epoch": 17.720822566752798,
"grad_norm": 0.3763260543346405,
"learning_rate": 0.0003344190117140164,
"loss": 3.0563,
"step": 65850
},
{
"epoch": 17.73428079242033,
"grad_norm": 0.4094148278236389,
"learning_rate": 0.0003342170459135586,
"loss": 3.0592,
"step": 65900
},
{
"epoch": 17.747739018087856,
"grad_norm": 0.38038286566734314,
"learning_rate": 0.0003340150801131008,
"loss": 3.0458,
"step": 65950
},
{
"epoch": 17.761197243755383,
"grad_norm": 0.40296798944473267,
"learning_rate": 0.00033381311431264303,
"loss": 3.0561,
"step": 66000
},
{
"epoch": 17.761197243755383,
"eval_accuracy": 0.39518316338792187,
"eval_loss": 3.2837977409362793,
"eval_runtime": 53.1573,
"eval_samples_per_second": 338.844,
"eval_steps_per_second": 21.182,
"step": 66000
},
{
"epoch": 17.77465546942291,
"grad_norm": 0.371357262134552,
"learning_rate": 0.00033361114851218523,
"loss": 3.0456,
"step": 66050
},
{
"epoch": 17.788113695090438,
"grad_norm": 0.4011745750904083,
"learning_rate": 0.0003334091827117274,
"loss": 3.0548,
"step": 66100
},
{
"epoch": 17.80157192075797,
"grad_norm": 0.3885413408279419,
"learning_rate": 0.0003332072169112696,
"loss": 3.0575,
"step": 66150
},
{
"epoch": 17.815030146425496,
"grad_norm": 0.40944671630859375,
"learning_rate": 0.0003330052511108118,
"loss": 3.0611,
"step": 66200
},
{
"epoch": 17.828488372093023,
"grad_norm": 0.38013756275177,
"learning_rate": 0.0003328032853103541,
"loss": 3.0549,
"step": 66250
},
{
"epoch": 17.84194659776055,
"grad_norm": 0.395384281873703,
"learning_rate": 0.0003326013195098963,
"loss": 3.056,
"step": 66300
},
{
"epoch": 17.855404823428078,
"grad_norm": 0.3956717550754547,
"learning_rate": 0.0003323993537094385,
"loss": 3.0669,
"step": 66350
},
{
"epoch": 17.86886304909561,
"grad_norm": 0.402349591255188,
"learning_rate": 0.00033219738790898077,
"loss": 3.0615,
"step": 66400
},
{
"epoch": 17.882321274763136,
"grad_norm": 0.4142409563064575,
"learning_rate": 0.00033199542210852296,
"loss": 3.0672,
"step": 66450
},
{
"epoch": 17.895779500430663,
"grad_norm": 0.414682537317276,
"learning_rate": 0.00033179345630806516,
"loss": 3.0628,
"step": 66500
},
{
"epoch": 17.90923772609819,
"grad_norm": 0.392534464597702,
"learning_rate": 0.00033159149050760735,
"loss": 3.0644,
"step": 66550
},
{
"epoch": 17.922695951765718,
"grad_norm": 0.41097918152809143,
"learning_rate": 0.00033138952470714955,
"loss": 3.0686,
"step": 66600
},
{
"epoch": 17.93615417743325,
"grad_norm": 0.407552570104599,
"learning_rate": 0.0003311875589066918,
"loss": 3.0715,
"step": 66650
},
{
"epoch": 17.949612403100776,
"grad_norm": 0.38662219047546387,
"learning_rate": 0.000330985593106234,
"loss": 3.0591,
"step": 66700
},
{
"epoch": 17.963070628768303,
"grad_norm": 0.41780713200569153,
"learning_rate": 0.0003307836273057762,
"loss": 3.058,
"step": 66750
},
{
"epoch": 17.97652885443583,
"grad_norm": 0.3943421244621277,
"learning_rate": 0.0003305816615053184,
"loss": 3.064,
"step": 66800
},
{
"epoch": 17.989987080103358,
"grad_norm": 0.37973111867904663,
"learning_rate": 0.0003303796957048606,
"loss": 3.065,
"step": 66850
},
{
"epoch": 18.003229974160206,
"grad_norm": 0.42041251063346863,
"learning_rate": 0.00033017772990440284,
"loss": 3.0416,
"step": 66900
},
{
"epoch": 18.016688199827733,
"grad_norm": 0.4095908999443054,
"learning_rate": 0.00032997576410394503,
"loss": 2.9723,
"step": 66950
},
{
"epoch": 18.030146425495264,
"grad_norm": 0.43125247955322266,
"learning_rate": 0.00032977379830348723,
"loss": 2.9748,
"step": 67000
},
{
"epoch": 18.030146425495264,
"eval_accuracy": 0.39475335642730897,
"eval_loss": 3.293449878692627,
"eval_runtime": 53.2,
"eval_samples_per_second": 338.572,
"eval_steps_per_second": 21.165,
"step": 67000
},
{
"epoch": 18.04360465116279,
"grad_norm": 0.4348003566265106,
"learning_rate": 0.0003295718325030294,
"loss": 2.9726,
"step": 67050
},
{
"epoch": 18.05706287683032,
"grad_norm": 0.4124617576599121,
"learning_rate": 0.0003293698667025716,
"loss": 2.9857,
"step": 67100
},
{
"epoch": 18.070521102497846,
"grad_norm": 0.4146093726158142,
"learning_rate": 0.0003291679009021139,
"loss": 2.9821,
"step": 67150
},
{
"epoch": 18.083979328165373,
"grad_norm": 0.41040223836898804,
"learning_rate": 0.0003289659351016561,
"loss": 2.9869,
"step": 67200
},
{
"epoch": 18.097437553832904,
"grad_norm": 0.4050770401954651,
"learning_rate": 0.0003287639693011983,
"loss": 2.9874,
"step": 67250
},
{
"epoch": 18.11089577950043,
"grad_norm": 0.39382511377334595,
"learning_rate": 0.0003285620035007405,
"loss": 3.0004,
"step": 67300
},
{
"epoch": 18.12435400516796,
"grad_norm": 0.42360246181488037,
"learning_rate": 0.00032836003770028276,
"loss": 2.9892,
"step": 67350
},
{
"epoch": 18.137812230835486,
"grad_norm": 0.46744874119758606,
"learning_rate": 0.00032815807189982496,
"loss": 2.9902,
"step": 67400
},
{
"epoch": 18.151270456503013,
"grad_norm": 0.4184926450252533,
"learning_rate": 0.00032795610609936715,
"loss": 2.9957,
"step": 67450
},
{
"epoch": 18.164728682170544,
"grad_norm": 0.4146018624305725,
"learning_rate": 0.00032775414029890935,
"loss": 2.9911,
"step": 67500
},
{
"epoch": 18.17818690783807,
"grad_norm": 0.40711694955825806,
"learning_rate": 0.00032755217449845155,
"loss": 3.0022,
"step": 67550
},
{
"epoch": 18.1916451335056,
"grad_norm": 0.3984438478946686,
"learning_rate": 0.0003273502086979938,
"loss": 3.0039,
"step": 67600
},
{
"epoch": 18.205103359173126,
"grad_norm": 0.40462562441825867,
"learning_rate": 0.000327148242897536,
"loss": 2.9994,
"step": 67650
},
{
"epoch": 18.218561584840653,
"grad_norm": 0.47321245074272156,
"learning_rate": 0.0003269462770970782,
"loss": 3.0075,
"step": 67700
},
{
"epoch": 18.232019810508184,
"grad_norm": 0.4014117419719696,
"learning_rate": 0.0003267443112966204,
"loss": 3.0112,
"step": 67750
},
{
"epoch": 18.24547803617571,
"grad_norm": 0.42581304907798767,
"learning_rate": 0.0003265423454961626,
"loss": 3.0134,
"step": 67800
},
{
"epoch": 18.25893626184324,
"grad_norm": 0.4282774031162262,
"learning_rate": 0.00032634037969570483,
"loss": 3.0109,
"step": 67850
},
{
"epoch": 18.272394487510766,
"grad_norm": 0.3891073167324066,
"learning_rate": 0.00032613841389524703,
"loss": 3.0121,
"step": 67900
},
{
"epoch": 18.285852713178294,
"grad_norm": 0.449763685464859,
"learning_rate": 0.0003259364480947892,
"loss": 2.9989,
"step": 67950
},
{
"epoch": 18.29931093884582,
"grad_norm": 0.4044979512691498,
"learning_rate": 0.0003257344822943314,
"loss": 3.004,
"step": 68000
},
{
"epoch": 18.29931093884582,
"eval_accuracy": 0.3946264569000097,
"eval_loss": 3.295100688934326,
"eval_runtime": 53.5137,
"eval_samples_per_second": 336.587,
"eval_steps_per_second": 21.041,
"step": 68000
},
{
"epoch": 18.31276916451335,
"grad_norm": 0.4079665243625641,
"learning_rate": 0.0003255325164938737,
"loss": 3.0164,
"step": 68050
},
{
"epoch": 18.32622739018088,
"grad_norm": 0.3965129852294922,
"learning_rate": 0.0003253305506934159,
"loss": 3.0175,
"step": 68100
},
{
"epoch": 18.339685615848406,
"grad_norm": 0.4107968509197235,
"learning_rate": 0.0003251285848929581,
"loss": 3.0203,
"step": 68150
},
{
"epoch": 18.353143841515934,
"grad_norm": 0.4172326326370239,
"learning_rate": 0.0003249266190925003,
"loss": 3.0209,
"step": 68200
},
{
"epoch": 18.36660206718346,
"grad_norm": 0.39596325159072876,
"learning_rate": 0.0003247246532920425,
"loss": 3.0203,
"step": 68250
},
{
"epoch": 18.38006029285099,
"grad_norm": 0.42070481181144714,
"learning_rate": 0.00032452268749158476,
"loss": 3.0214,
"step": 68300
},
{
"epoch": 18.39351851851852,
"grad_norm": 0.4123518466949463,
"learning_rate": 0.00032432072169112696,
"loss": 3.0302,
"step": 68350
},
{
"epoch": 18.406976744186046,
"grad_norm": 0.46176663041114807,
"learning_rate": 0.00032411875589066915,
"loss": 3.0233,
"step": 68400
},
{
"epoch": 18.420434969853574,
"grad_norm": 0.42506712675094604,
"learning_rate": 0.00032391679009021135,
"loss": 3.0217,
"step": 68450
},
{
"epoch": 18.4338931955211,
"grad_norm": 0.38221848011016846,
"learning_rate": 0.0003237148242897536,
"loss": 3.0227,
"step": 68500
},
{
"epoch": 18.447351421188632,
"grad_norm": 0.43396782875061035,
"learning_rate": 0.0003235128584892958,
"loss": 3.0201,
"step": 68550
},
{
"epoch": 18.46080964685616,
"grad_norm": 0.43768152594566345,
"learning_rate": 0.000323310892688838,
"loss": 3.0195,
"step": 68600
},
{
"epoch": 18.474267872523686,
"grad_norm": 0.40827372670173645,
"learning_rate": 0.0003231089268883802,
"loss": 3.0232,
"step": 68650
},
{
"epoch": 18.487726098191214,
"grad_norm": 0.4054965078830719,
"learning_rate": 0.0003229069610879224,
"loss": 3.0244,
"step": 68700
},
{
"epoch": 18.50118432385874,
"grad_norm": 0.40097054839134216,
"learning_rate": 0.00032270499528746463,
"loss": 3.0317,
"step": 68750
},
{
"epoch": 18.514642549526272,
"grad_norm": 0.41187137365341187,
"learning_rate": 0.00032250302948700683,
"loss": 3.0275,
"step": 68800
},
{
"epoch": 18.5281007751938,
"grad_norm": 0.42199191451072693,
"learning_rate": 0.000322301063686549,
"loss": 3.0334,
"step": 68850
},
{
"epoch": 18.541559000861326,
"grad_norm": 0.38072946667671204,
"learning_rate": 0.0003220990978860912,
"loss": 3.0353,
"step": 68900
},
{
"epoch": 18.555017226528854,
"grad_norm": 0.4191347360610962,
"learning_rate": 0.0003218971320856335,
"loss": 3.0401,
"step": 68950
},
{
"epoch": 18.56847545219638,
"grad_norm": 0.413581907749176,
"learning_rate": 0.0003216951662851757,
"loss": 3.0315,
"step": 69000
},
{
"epoch": 18.56847545219638,
"eval_accuracy": 0.3948620032828734,
"eval_loss": 3.2892377376556396,
"eval_runtime": 53.2832,
"eval_samples_per_second": 338.043,
"eval_steps_per_second": 21.132,
"step": 69000
},
{
"epoch": 18.581933677863912,
"grad_norm": 0.431535005569458,
"learning_rate": 0.0003214932004847179,
"loss": 3.0412,
"step": 69050
},
{
"epoch": 18.59539190353144,
"grad_norm": 0.4194125831127167,
"learning_rate": 0.0003212912346842601,
"loss": 3.0391,
"step": 69100
},
{
"epoch": 18.608850129198967,
"grad_norm": 0.4209112524986267,
"learning_rate": 0.0003210892688838023,
"loss": 3.0389,
"step": 69150
},
{
"epoch": 18.622308354866494,
"grad_norm": 0.41800597310066223,
"learning_rate": 0.00032088730308334456,
"loss": 3.0292,
"step": 69200
},
{
"epoch": 18.63576658053402,
"grad_norm": 0.39001700282096863,
"learning_rate": 0.00032068533728288676,
"loss": 3.0326,
"step": 69250
},
{
"epoch": 18.649224806201552,
"grad_norm": 0.38095352053642273,
"learning_rate": 0.00032048337148242895,
"loss": 3.0305,
"step": 69300
},
{
"epoch": 18.66268303186908,
"grad_norm": 0.40673938393592834,
"learning_rate": 0.00032028140568197115,
"loss": 3.0367,
"step": 69350
},
{
"epoch": 18.676141257536607,
"grad_norm": 0.40681397914886475,
"learning_rate": 0.00032007943988151334,
"loss": 3.0247,
"step": 69400
},
{
"epoch": 18.689599483204134,
"grad_norm": 0.40793168544769287,
"learning_rate": 0.0003198774740810556,
"loss": 3.0374,
"step": 69450
},
{
"epoch": 18.70305770887166,
"grad_norm": 0.4070163667201996,
"learning_rate": 0.0003196755082805978,
"loss": 3.0361,
"step": 69500
},
{
"epoch": 18.716515934539192,
"grad_norm": 0.40486323833465576,
"learning_rate": 0.00031947354248014,
"loss": 3.0403,
"step": 69550
},
{
"epoch": 18.72997416020672,
"grad_norm": 0.4136981666088104,
"learning_rate": 0.0003192715766796822,
"loss": 3.0447,
"step": 69600
},
{
"epoch": 18.743432385874247,
"grad_norm": 0.41730913519859314,
"learning_rate": 0.0003190696108792244,
"loss": 3.0411,
"step": 69650
},
{
"epoch": 18.756890611541774,
"grad_norm": 0.4055947959423065,
"learning_rate": 0.00031886764507876663,
"loss": 3.0463,
"step": 69700
},
{
"epoch": 18.7703488372093,
"grad_norm": 0.4054621756076813,
"learning_rate": 0.0003186656792783088,
"loss": 3.0443,
"step": 69750
},
{
"epoch": 18.783807062876832,
"grad_norm": 0.42214059829711914,
"learning_rate": 0.000318463713477851,
"loss": 3.0503,
"step": 69800
},
{
"epoch": 18.79726528854436,
"grad_norm": 0.43466436862945557,
"learning_rate": 0.00031826174767739327,
"loss": 3.0506,
"step": 69850
},
{
"epoch": 18.810723514211887,
"grad_norm": 0.3966998755931854,
"learning_rate": 0.0003180597818769355,
"loss": 3.0378,
"step": 69900
},
{
"epoch": 18.824181739879414,
"grad_norm": 0.422006219625473,
"learning_rate": 0.0003178578160764777,
"loss": 3.0567,
"step": 69950
},
{
"epoch": 18.83763996554694,
"grad_norm": 0.4191000759601593,
"learning_rate": 0.0003176558502760199,
"loss": 3.0424,
"step": 70000
},
{
"epoch": 18.83763996554694,
"eval_accuracy": 0.3954330511557201,
"eval_loss": 3.278806209564209,
"eval_runtime": 53.1978,
"eval_samples_per_second": 338.585,
"eval_steps_per_second": 21.166,
"step": 70000
},
{
"epoch": 18.85109819121447,
"grad_norm": 0.4105396866798401,
"learning_rate": 0.0003174538844755621,
"loss": 3.0357,
"step": 70050
},
{
"epoch": 18.864556416882,
"grad_norm": 0.4191843271255493,
"learning_rate": 0.0003172519186751043,
"loss": 3.0515,
"step": 70100
},
{
"epoch": 18.878014642549527,
"grad_norm": 0.3951943516731262,
"learning_rate": 0.00031704995287464656,
"loss": 3.0501,
"step": 70150
},
{
"epoch": 18.891472868217054,
"grad_norm": 0.3969205915927887,
"learning_rate": 0.00031684798707418875,
"loss": 3.0475,
"step": 70200
},
{
"epoch": 18.90493109388458,
"grad_norm": 0.4015067219734192,
"learning_rate": 0.00031664602127373095,
"loss": 3.0622,
"step": 70250
},
{
"epoch": 18.91838931955211,
"grad_norm": 0.39716896414756775,
"learning_rate": 0.00031644405547327315,
"loss": 3.0546,
"step": 70300
},
{
"epoch": 18.93184754521964,
"grad_norm": 0.42040297389030457,
"learning_rate": 0.00031624208967281534,
"loss": 3.0566,
"step": 70350
},
{
"epoch": 18.945305770887167,
"grad_norm": 0.397980272769928,
"learning_rate": 0.0003160401238723576,
"loss": 3.0437,
"step": 70400
},
{
"epoch": 18.958763996554694,
"grad_norm": 0.42787206172943115,
"learning_rate": 0.0003158381580718998,
"loss": 3.0559,
"step": 70450
},
{
"epoch": 18.97222222222222,
"grad_norm": 0.4149838089942932,
"learning_rate": 0.000315636192271442,
"loss": 3.0584,
"step": 70500
},
{
"epoch": 18.98568044788975,
"grad_norm": 0.4251541197299957,
"learning_rate": 0.0003154342264709842,
"loss": 3.0621,
"step": 70550
},
{
"epoch": 18.99913867355728,
"grad_norm": 0.41713324189186096,
"learning_rate": 0.00031523226067052643,
"loss": 3.0395,
"step": 70600
},
{
"epoch": 19.012381567614124,
"grad_norm": 0.43764570355415344,
"learning_rate": 0.0003150302948700686,
"loss": 2.9625,
"step": 70650
},
{
"epoch": 19.025839793281655,
"grad_norm": 0.4346773326396942,
"learning_rate": 0.0003148283290696108,
"loss": 2.9584,
"step": 70700
},
{
"epoch": 19.039298018949182,
"grad_norm": 0.4352344274520874,
"learning_rate": 0.0003146263632691531,
"loss": 2.9643,
"step": 70750
},
{
"epoch": 19.05275624461671,
"grad_norm": 0.4444960951805115,
"learning_rate": 0.0003144243974686953,
"loss": 2.9676,
"step": 70800
},
{
"epoch": 19.066214470284237,
"grad_norm": 0.41851648688316345,
"learning_rate": 0.0003142224316682375,
"loss": 2.9715,
"step": 70850
},
{
"epoch": 19.079672695951764,
"grad_norm": 0.4398450553417206,
"learning_rate": 0.0003140204658677797,
"loss": 2.9952,
"step": 70900
},
{
"epoch": 19.093130921619295,
"grad_norm": 0.39954739809036255,
"learning_rate": 0.0003138185000673219,
"loss": 2.9676,
"step": 70950
},
{
"epoch": 19.106589147286822,
"grad_norm": 0.42698994278907776,
"learning_rate": 0.0003136165342668641,
"loss": 2.9799,
"step": 71000
},
{
"epoch": 19.106589147286822,
"eval_accuracy": 0.3946730664010468,
"eval_loss": 3.2944045066833496,
"eval_runtime": 53.1951,
"eval_samples_per_second": 338.602,
"eval_steps_per_second": 21.167,
"step": 71000
},
{
"epoch": 19.12004737295435,
"grad_norm": 0.4176847040653229,
"learning_rate": 0.00031341456846640636,
"loss": 2.9882,
"step": 71050
},
{
"epoch": 19.133505598621877,
"grad_norm": 0.43521398305892944,
"learning_rate": 0.00031321260266594855,
"loss": 2.9847,
"step": 71100
},
{
"epoch": 19.146963824289404,
"grad_norm": 0.4214783012866974,
"learning_rate": 0.00031301063686549075,
"loss": 2.9982,
"step": 71150
},
{
"epoch": 19.160422049956935,
"grad_norm": 0.4213625192642212,
"learning_rate": 0.00031280867106503295,
"loss": 2.9853,
"step": 71200
},
{
"epoch": 19.173880275624462,
"grad_norm": 0.40806859731674194,
"learning_rate": 0.00031260670526457514,
"loss": 2.9855,
"step": 71250
},
{
"epoch": 19.18733850129199,
"grad_norm": 0.4534761905670166,
"learning_rate": 0.0003124047394641174,
"loss": 2.9845,
"step": 71300
},
{
"epoch": 19.200796726959517,
"grad_norm": 0.4072936773300171,
"learning_rate": 0.0003122027736636596,
"loss": 2.9865,
"step": 71350
},
{
"epoch": 19.214254952627044,
"grad_norm": 0.40868863463401794,
"learning_rate": 0.0003120008078632018,
"loss": 2.9928,
"step": 71400
},
{
"epoch": 19.227713178294575,
"grad_norm": 0.4284927546977997,
"learning_rate": 0.000311798842062744,
"loss": 3.0037,
"step": 71450
},
{
"epoch": 19.241171403962102,
"grad_norm": 0.4234805405139923,
"learning_rate": 0.0003115968762622862,
"loss": 2.9812,
"step": 71500
},
{
"epoch": 19.25462962962963,
"grad_norm": 0.40032583475112915,
"learning_rate": 0.00031139491046182843,
"loss": 3.0,
"step": 71550
},
{
"epoch": 19.268087855297157,
"grad_norm": 0.4138863980770111,
"learning_rate": 0.0003111929446613707,
"loss": 2.9892,
"step": 71600
},
{
"epoch": 19.281546080964684,
"grad_norm": 0.4290546476840973,
"learning_rate": 0.0003109909788609129,
"loss": 3.0028,
"step": 71650
},
{
"epoch": 19.295004306632215,
"grad_norm": 0.4247094690799713,
"learning_rate": 0.00031078901306045507,
"loss": 2.9972,
"step": 71700
},
{
"epoch": 19.308462532299743,
"grad_norm": 0.4409559965133667,
"learning_rate": 0.0003105870472599973,
"loss": 3.0053,
"step": 71750
},
{
"epoch": 19.32192075796727,
"grad_norm": 0.4488597810268402,
"learning_rate": 0.0003103850814595395,
"loss": 2.9962,
"step": 71800
},
{
"epoch": 19.335378983634797,
"grad_norm": 0.4389813244342804,
"learning_rate": 0.0003101831156590817,
"loss": 3.0027,
"step": 71850
},
{
"epoch": 19.348837209302324,
"grad_norm": 0.40503010153770447,
"learning_rate": 0.0003099811498586239,
"loss": 3.0004,
"step": 71900
},
{
"epoch": 19.362295434969855,
"grad_norm": 0.44304975867271423,
"learning_rate": 0.0003097791840581661,
"loss": 3.0097,
"step": 71950
},
{
"epoch": 19.375753660637383,
"grad_norm": 0.4481963813304901,
"learning_rate": 0.00030957721825770836,
"loss": 2.999,
"step": 72000
},
{
"epoch": 19.375753660637383,
"eval_accuracy": 0.3951246027327726,
"eval_loss": 3.290699005126953,
"eval_runtime": 53.1572,
"eval_samples_per_second": 338.844,
"eval_steps_per_second": 21.182,
"step": 72000
},
{
"epoch": 19.38921188630491,
"grad_norm": 0.4432748258113861,
"learning_rate": 0.00030937525245725055,
"loss": 3.0115,
"step": 72050
},
{
"epoch": 19.402670111972437,
"grad_norm": 0.4061640799045563,
"learning_rate": 0.00030917328665679275,
"loss": 3.0055,
"step": 72100
},
{
"epoch": 19.416128337639964,
"grad_norm": 0.4457055926322937,
"learning_rate": 0.00030897132085633494,
"loss": 3.0143,
"step": 72150
},
{
"epoch": 19.429586563307495,
"grad_norm": 0.44451671838760376,
"learning_rate": 0.00030876935505587714,
"loss": 3.0057,
"step": 72200
},
{
"epoch": 19.443044788975023,
"grad_norm": 0.41940534114837646,
"learning_rate": 0.0003085673892554194,
"loss": 3.0211,
"step": 72250
},
{
"epoch": 19.45650301464255,
"grad_norm": 0.40035170316696167,
"learning_rate": 0.0003083654234549616,
"loss": 3.0135,
"step": 72300
},
{
"epoch": 19.469961240310077,
"grad_norm": 0.42814239859580994,
"learning_rate": 0.0003081634576545038,
"loss": 3.0146,
"step": 72350
},
{
"epoch": 19.483419465977605,
"grad_norm": 0.42938923835754395,
"learning_rate": 0.000307961491854046,
"loss": 3.0184,
"step": 72400
},
{
"epoch": 19.496877691645132,
"grad_norm": 0.40963107347488403,
"learning_rate": 0.00030775952605358823,
"loss": 3.0071,
"step": 72450
},
{
"epoch": 19.510335917312663,
"grad_norm": 0.4201676845550537,
"learning_rate": 0.0003075575602531305,
"loss": 3.0049,
"step": 72500
},
{
"epoch": 19.52379414298019,
"grad_norm": 0.40009865164756775,
"learning_rate": 0.0003073555944526727,
"loss": 3.0141,
"step": 72550
},
{
"epoch": 19.537252368647717,
"grad_norm": 0.4132072925567627,
"learning_rate": 0.00030715362865221487,
"loss": 3.0243,
"step": 72600
},
{
"epoch": 19.550710594315245,
"grad_norm": 0.4008041024208069,
"learning_rate": 0.0003069516628517571,
"loss": 3.0209,
"step": 72650
},
{
"epoch": 19.564168819982772,
"grad_norm": 0.4381272792816162,
"learning_rate": 0.0003067496970512993,
"loss": 3.0309,
"step": 72700
},
{
"epoch": 19.577627045650303,
"grad_norm": 0.4061357378959656,
"learning_rate": 0.0003065477312508415,
"loss": 3.0198,
"step": 72750
},
{
"epoch": 19.59108527131783,
"grad_norm": 0.42047545313835144,
"learning_rate": 0.0003063457654503837,
"loss": 3.0306,
"step": 72800
},
{
"epoch": 19.604543496985357,
"grad_norm": 0.411575049161911,
"learning_rate": 0.0003061437996499259,
"loss": 3.0281,
"step": 72850
},
{
"epoch": 19.618001722652885,
"grad_norm": 0.42404329776763916,
"learning_rate": 0.00030594183384946816,
"loss": 3.0326,
"step": 72900
},
{
"epoch": 19.631459948320412,
"grad_norm": 0.42757225036621094,
"learning_rate": 0.00030573986804901035,
"loss": 3.0259,
"step": 72950
},
{
"epoch": 19.644918173987943,
"grad_norm": 0.3978443741798401,
"learning_rate": 0.00030553790224855255,
"loss": 3.0285,
"step": 73000
},
{
"epoch": 19.644918173987943,
"eval_accuracy": 0.3953907875289055,
"eval_loss": 3.284182548522949,
"eval_runtime": 53.2956,
"eval_samples_per_second": 337.964,
"eval_steps_per_second": 21.127,
"step": 73000
},
{
"epoch": 19.65837639965547,
"grad_norm": 0.3940031826496124,
"learning_rate": 0.00030533593644809474,
"loss": 3.0237,
"step": 73050
},
{
"epoch": 19.671834625322997,
"grad_norm": 0.44861793518066406,
"learning_rate": 0.00030513397064763694,
"loss": 3.0246,
"step": 73100
},
{
"epoch": 19.685292850990525,
"grad_norm": 0.42165327072143555,
"learning_rate": 0.0003049320048471792,
"loss": 3.0414,
"step": 73150
},
{
"epoch": 19.698751076658052,
"grad_norm": 0.391513854265213,
"learning_rate": 0.0003047300390467214,
"loss": 3.0232,
"step": 73200
},
{
"epoch": 19.712209302325583,
"grad_norm": 0.4060502052307129,
"learning_rate": 0.0003045280732462636,
"loss": 3.0291,
"step": 73250
},
{
"epoch": 19.72566752799311,
"grad_norm": 0.39743009209632874,
"learning_rate": 0.0003043261074458058,
"loss": 3.0313,
"step": 73300
},
{
"epoch": 19.739125753660637,
"grad_norm": 0.41127610206604004,
"learning_rate": 0.000304124141645348,
"loss": 3.0281,
"step": 73350
},
{
"epoch": 19.752583979328165,
"grad_norm": 0.398449569940567,
"learning_rate": 0.0003039221758448903,
"loss": 3.0166,
"step": 73400
},
{
"epoch": 19.766042204995692,
"grad_norm": 0.4216623604297638,
"learning_rate": 0.0003037202100444325,
"loss": 3.0423,
"step": 73450
},
{
"epoch": 19.779500430663223,
"grad_norm": 0.4148426949977875,
"learning_rate": 0.00030351824424397467,
"loss": 3.0202,
"step": 73500
},
{
"epoch": 19.79295865633075,
"grad_norm": 0.44049158692359924,
"learning_rate": 0.00030331627844351687,
"loss": 3.0349,
"step": 73550
},
{
"epoch": 19.806416881998278,
"grad_norm": 0.4221228361129761,
"learning_rate": 0.0003031143126430591,
"loss": 3.0326,
"step": 73600
},
{
"epoch": 19.819875107665805,
"grad_norm": 0.41475948691368103,
"learning_rate": 0.0003029123468426013,
"loss": 3.035,
"step": 73650
},
{
"epoch": 19.833333333333332,
"grad_norm": 0.41580721735954285,
"learning_rate": 0.0003027103810421435,
"loss": 3.0412,
"step": 73700
},
{
"epoch": 19.846791559000863,
"grad_norm": 0.4306314289569855,
"learning_rate": 0.0003025084152416857,
"loss": 3.0312,
"step": 73750
},
{
"epoch": 19.86024978466839,
"grad_norm": 0.41465747356414795,
"learning_rate": 0.0003023064494412279,
"loss": 3.0449,
"step": 73800
},
{
"epoch": 19.873708010335918,
"grad_norm": 0.427436888217926,
"learning_rate": 0.00030210448364077015,
"loss": 3.0337,
"step": 73850
},
{
"epoch": 19.887166236003445,
"grad_norm": 0.4142983555793762,
"learning_rate": 0.00030190251784031235,
"loss": 3.0446,
"step": 73900
},
{
"epoch": 19.900624461670972,
"grad_norm": 0.42197173833847046,
"learning_rate": 0.00030170055203985454,
"loss": 3.0457,
"step": 73950
},
{
"epoch": 19.9140826873385,
"grad_norm": 0.41774898767471313,
"learning_rate": 0.00030149858623939674,
"loss": 3.0365,
"step": 74000
},
{
"epoch": 19.9140826873385,
"eval_accuracy": 0.3958141843250401,
"eval_loss": 3.279388427734375,
"eval_runtime": 53.2046,
"eval_samples_per_second": 338.542,
"eval_steps_per_second": 21.164,
"step": 74000
},
{
"epoch": 19.92754091300603,
"grad_norm": 0.47798389196395874,
"learning_rate": 0.00030129662043893894,
"loss": 3.0397,
"step": 74050
},
{
"epoch": 19.940999138673558,
"grad_norm": 0.44100531935691833,
"learning_rate": 0.0003010946546384812,
"loss": 3.046,
"step": 74100
},
{
"epoch": 19.954457364341085,
"grad_norm": 0.42226624488830566,
"learning_rate": 0.0003008926888380234,
"loss": 3.0389,
"step": 74150
},
{
"epoch": 19.967915590008612,
"grad_norm": 0.3779127597808838,
"learning_rate": 0.0003006907230375656,
"loss": 3.0381,
"step": 74200
},
{
"epoch": 19.98137381567614,
"grad_norm": 0.4114772379398346,
"learning_rate": 0.0003004887572371078,
"loss": 3.0382,
"step": 74250
},
{
"epoch": 19.99483204134367,
"grad_norm": 0.4274559020996094,
"learning_rate": 0.0003002867914366501,
"loss": 3.0433,
"step": 74300
},
{
"epoch": 20.00807493540052,
"grad_norm": 0.40736275911331177,
"learning_rate": 0.0003000848256361923,
"loss": 2.9855,
"step": 74350
},
{
"epoch": 20.021533161068046,
"grad_norm": 0.4333912134170532,
"learning_rate": 0.00029988285983573447,
"loss": 2.939,
"step": 74400
},
{
"epoch": 20.034991386735573,
"grad_norm": 0.42243754863739014,
"learning_rate": 0.00029968089403527667,
"loss": 2.9527,
"step": 74450
},
{
"epoch": 20.0484496124031,
"grad_norm": 0.43641528487205505,
"learning_rate": 0.0002994789282348189,
"loss": 2.962,
"step": 74500
},
{
"epoch": 20.061907838070628,
"grad_norm": 0.42667356133461,
"learning_rate": 0.0002992769624343611,
"loss": 2.9601,
"step": 74550
},
{
"epoch": 20.075366063738155,
"grad_norm": 0.4108620285987854,
"learning_rate": 0.0002990749966339033,
"loss": 2.9652,
"step": 74600
},
{
"epoch": 20.088824289405686,
"grad_norm": 0.4250992238521576,
"learning_rate": 0.0002988730308334455,
"loss": 2.9587,
"step": 74650
},
{
"epoch": 20.102282515073213,
"grad_norm": 0.41763296723365784,
"learning_rate": 0.0002986710650329877,
"loss": 2.9691,
"step": 74700
},
{
"epoch": 20.11574074074074,
"grad_norm": 0.4523560106754303,
"learning_rate": 0.00029846909923252995,
"loss": 2.971,
"step": 74750
},
{
"epoch": 20.129198966408268,
"grad_norm": 0.4249483048915863,
"learning_rate": 0.00029826713343207215,
"loss": 2.9687,
"step": 74800
},
{
"epoch": 20.142657192075795,
"grad_norm": 0.4228469133377075,
"learning_rate": 0.00029806516763161435,
"loss": 2.9815,
"step": 74850
},
{
"epoch": 20.156115417743326,
"grad_norm": 0.4154616892337799,
"learning_rate": 0.00029786320183115654,
"loss": 2.9821,
"step": 74900
},
{
"epoch": 20.169573643410853,
"grad_norm": 0.43550071120262146,
"learning_rate": 0.0002976612360306988,
"loss": 2.9639,
"step": 74950
},
{
"epoch": 20.18303186907838,
"grad_norm": 0.415121853351593,
"learning_rate": 0.000297459270230241,
"loss": 2.979,
"step": 75000
},
{
"epoch": 20.18303186907838,
"eval_accuracy": 0.39531125803063233,
"eval_loss": 3.2909493446350098,
"eval_runtime": 53.2389,
"eval_samples_per_second": 338.324,
"eval_steps_per_second": 21.15,
"step": 75000
},
{
"epoch": 20.196490094745908,
"grad_norm": 0.4105370044708252,
"learning_rate": 0.0002972573044297832,
"loss": 2.9781,
"step": 75050
},
{
"epoch": 20.209948320413435,
"grad_norm": 0.4527175724506378,
"learning_rate": 0.00029705533862932543,
"loss": 2.9852,
"step": 75100
},
{
"epoch": 20.223406546080966,
"grad_norm": 0.40570807456970215,
"learning_rate": 0.00029685337282886763,
"loss": 2.9897,
"step": 75150
},
{
"epoch": 20.236864771748493,
"grad_norm": 0.43341004848480225,
"learning_rate": 0.0002966514070284098,
"loss": 2.9861,
"step": 75200
},
{
"epoch": 20.25032299741602,
"grad_norm": 0.44538331031799316,
"learning_rate": 0.000296449441227952,
"loss": 2.99,
"step": 75250
},
{
"epoch": 20.263781223083548,
"grad_norm": 0.40199220180511475,
"learning_rate": 0.0002962474754274942,
"loss": 2.9836,
"step": 75300
},
{
"epoch": 20.277239448751075,
"grad_norm": 0.4486052989959717,
"learning_rate": 0.00029604550962703647,
"loss": 2.9899,
"step": 75350
},
{
"epoch": 20.290697674418606,
"grad_norm": 0.44760704040527344,
"learning_rate": 0.00029584354382657867,
"loss": 2.9887,
"step": 75400
},
{
"epoch": 20.304155900086133,
"grad_norm": 0.42043086886405945,
"learning_rate": 0.0002956415780261209,
"loss": 2.9937,
"step": 75450
},
{
"epoch": 20.31761412575366,
"grad_norm": 0.43127012252807617,
"learning_rate": 0.0002954396122256631,
"loss": 2.9947,
"step": 75500
},
{
"epoch": 20.331072351421188,
"grad_norm": 0.4210168719291687,
"learning_rate": 0.0002952376464252053,
"loss": 2.9901,
"step": 75550
},
{
"epoch": 20.344530577088715,
"grad_norm": 0.42815783619880676,
"learning_rate": 0.0002950356806247475,
"loss": 3.0006,
"step": 75600
},
{
"epoch": 20.357988802756246,
"grad_norm": 0.4324301481246948,
"learning_rate": 0.0002948337148242897,
"loss": 2.9946,
"step": 75650
},
{
"epoch": 20.371447028423773,
"grad_norm": 0.40492090582847595,
"learning_rate": 0.00029463174902383195,
"loss": 2.9957,
"step": 75700
},
{
"epoch": 20.3849052540913,
"grad_norm": 0.41066285967826843,
"learning_rate": 0.00029442978322337415,
"loss": 2.9995,
"step": 75750
},
{
"epoch": 20.398363479758828,
"grad_norm": 0.45168185234069824,
"learning_rate": 0.00029422781742291634,
"loss": 2.993,
"step": 75800
},
{
"epoch": 20.411821705426355,
"grad_norm": 0.42091086506843567,
"learning_rate": 0.0002940258516224586,
"loss": 3.0148,
"step": 75850
},
{
"epoch": 20.425279931093886,
"grad_norm": 0.4229574203491211,
"learning_rate": 0.0002938238858220008,
"loss": 2.996,
"step": 75900
},
{
"epoch": 20.438738156761413,
"grad_norm": 0.4290357232093811,
"learning_rate": 0.000293621920021543,
"loss": 3.0044,
"step": 75950
},
{
"epoch": 20.45219638242894,
"grad_norm": 0.4104298949241638,
"learning_rate": 0.0002934199542210852,
"loss": 3.0131,
"step": 76000
},
{
"epoch": 20.45219638242894,
"eval_accuracy": 0.39542207782330807,
"eval_loss": 3.2870841026306152,
"eval_runtime": 53.2071,
"eval_samples_per_second": 338.526,
"eval_steps_per_second": 21.163,
"step": 76000
},
{
"epoch": 20.465654608096468,
"grad_norm": 0.43837058544158936,
"learning_rate": 0.00029321798842062743,
"loss": 3.0129,
"step": 76050
},
{
"epoch": 20.479112833763995,
"grad_norm": 0.4409915804862976,
"learning_rate": 0.00029301602262016963,
"loss": 2.9995,
"step": 76100
},
{
"epoch": 20.492571059431526,
"grad_norm": 0.4528588056564331,
"learning_rate": 0.0002928140568197118,
"loss": 3.0076,
"step": 76150
},
{
"epoch": 20.506029285099054,
"grad_norm": 0.41055387258529663,
"learning_rate": 0.000292612091019254,
"loss": 2.9977,
"step": 76200
},
{
"epoch": 20.51948751076658,
"grad_norm": 0.42017340660095215,
"learning_rate": 0.0002924101252187962,
"loss": 3.0144,
"step": 76250
},
{
"epoch": 20.532945736434108,
"grad_norm": 0.45756611227989197,
"learning_rate": 0.00029220815941833847,
"loss": 2.9964,
"step": 76300
},
{
"epoch": 20.546403962101635,
"grad_norm": 0.42427578568458557,
"learning_rate": 0.0002920061936178807,
"loss": 3.0133,
"step": 76350
},
{
"epoch": 20.559862187769163,
"grad_norm": 0.4466415345668793,
"learning_rate": 0.0002918042278174229,
"loss": 3.0102,
"step": 76400
},
{
"epoch": 20.573320413436694,
"grad_norm": 0.41783392429351807,
"learning_rate": 0.0002916022620169651,
"loss": 3.0047,
"step": 76450
},
{
"epoch": 20.58677863910422,
"grad_norm": 0.42117050290107727,
"learning_rate": 0.0002914002962165073,
"loss": 3.0129,
"step": 76500
},
{
"epoch": 20.600236864771748,
"grad_norm": 0.4089644253253937,
"learning_rate": 0.0002911983304160495,
"loss": 3.0128,
"step": 76550
},
{
"epoch": 20.613695090439276,
"grad_norm": 0.4382970333099365,
"learning_rate": 0.00029099636461559175,
"loss": 3.0158,
"step": 76600
},
{
"epoch": 20.627153316106803,
"grad_norm": 0.40870681405067444,
"learning_rate": 0.00029079439881513395,
"loss": 3.0157,
"step": 76650
},
{
"epoch": 20.640611541774334,
"grad_norm": 0.43205249309539795,
"learning_rate": 0.00029059243301467614,
"loss": 3.0142,
"step": 76700
},
{
"epoch": 20.65406976744186,
"grad_norm": 0.41538992524147034,
"learning_rate": 0.0002903904672142184,
"loss": 3.0206,
"step": 76750
},
{
"epoch": 20.66752799310939,
"grad_norm": 0.4724888801574707,
"learning_rate": 0.0002901885014137606,
"loss": 3.0077,
"step": 76800
},
{
"epoch": 20.680986218776916,
"grad_norm": 0.4144582450389862,
"learning_rate": 0.0002899865356133028,
"loss": 3.0272,
"step": 76850
},
{
"epoch": 20.694444444444443,
"grad_norm": 0.41652971506118774,
"learning_rate": 0.000289784569812845,
"loss": 3.0097,
"step": 76900
},
{
"epoch": 20.707902670111974,
"grad_norm": 0.43358200788497925,
"learning_rate": 0.00028958260401238723,
"loss": 3.0187,
"step": 76950
},
{
"epoch": 20.7213608957795,
"grad_norm": 0.41195759177207947,
"learning_rate": 0.00028938063821192943,
"loss": 3.0201,
"step": 77000
},
{
"epoch": 20.7213608957795,
"eval_accuracy": 0.39608623605137344,
"eval_loss": 3.2799088954925537,
"eval_runtime": 53.2969,
"eval_samples_per_second": 337.956,
"eval_steps_per_second": 21.127,
"step": 77000
},
{
"epoch": 20.73481912144703,
"grad_norm": 0.44272324442863464,
"learning_rate": 0.0002891786724114716,
"loss": 3.0161,
"step": 77050
},
{
"epoch": 20.748277347114556,
"grad_norm": 0.424826979637146,
"learning_rate": 0.0002889767066110138,
"loss": 3.0197,
"step": 77100
},
{
"epoch": 20.761735572782083,
"grad_norm": 0.4209699034690857,
"learning_rate": 0.000288774740810556,
"loss": 3.0277,
"step": 77150
},
{
"epoch": 20.775193798449614,
"grad_norm": 0.41250577569007874,
"learning_rate": 0.00028857277501009827,
"loss": 3.0296,
"step": 77200
},
{
"epoch": 20.78865202411714,
"grad_norm": 0.42251619696617126,
"learning_rate": 0.00028837080920964046,
"loss": 3.0234,
"step": 77250
},
{
"epoch": 20.80211024978467,
"grad_norm": 0.4229857325553894,
"learning_rate": 0.0002881688434091827,
"loss": 3.0211,
"step": 77300
},
{
"epoch": 20.815568475452196,
"grad_norm": 0.42098045349121094,
"learning_rate": 0.0002879668776087249,
"loss": 3.0308,
"step": 77350
},
{
"epoch": 20.829026701119723,
"grad_norm": 0.4067869782447815,
"learning_rate": 0.0002877649118082671,
"loss": 3.0273,
"step": 77400
},
{
"epoch": 20.842484926787254,
"grad_norm": 0.436855286359787,
"learning_rate": 0.0002875629460078093,
"loss": 3.0315,
"step": 77450
},
{
"epoch": 20.85594315245478,
"grad_norm": 0.4486081004142761,
"learning_rate": 0.0002873609802073515,
"loss": 3.0115,
"step": 77500
},
{
"epoch": 20.86940137812231,
"grad_norm": 0.40413883328437805,
"learning_rate": 0.00028715901440689375,
"loss": 3.0183,
"step": 77550
},
{
"epoch": 20.882859603789836,
"grad_norm": 0.4487050771713257,
"learning_rate": 0.00028695704860643594,
"loss": 3.0208,
"step": 77600
},
{
"epoch": 20.896317829457363,
"grad_norm": 0.43203026056289673,
"learning_rate": 0.0002867550828059782,
"loss": 3.0223,
"step": 77650
},
{
"epoch": 20.909776055124894,
"grad_norm": 0.4328913986682892,
"learning_rate": 0.0002865531170055204,
"loss": 3.0271,
"step": 77700
},
{
"epoch": 20.92323428079242,
"grad_norm": 0.39035165309906006,
"learning_rate": 0.0002863511512050626,
"loss": 3.031,
"step": 77750
},
{
"epoch": 20.93669250645995,
"grad_norm": 0.404269814491272,
"learning_rate": 0.0002861491854046048,
"loss": 3.0262,
"step": 77800
},
{
"epoch": 20.950150732127476,
"grad_norm": 0.40351414680480957,
"learning_rate": 0.000285947219604147,
"loss": 3.0221,
"step": 77850
},
{
"epoch": 20.963608957795003,
"grad_norm": 0.4256989657878876,
"learning_rate": 0.00028574525380368923,
"loss": 3.0303,
"step": 77900
},
{
"epoch": 20.977067183462534,
"grad_norm": 0.4155173599720001,
"learning_rate": 0.0002855432880032314,
"loss": 3.0339,
"step": 77950
},
{
"epoch": 20.99052540913006,
"grad_norm": 0.42269203066825867,
"learning_rate": 0.0002853413222027736,
"loss": 3.0228,
"step": 78000
},
{
"epoch": 20.99052540913006,
"eval_accuracy": 0.39598541176940966,
"eval_loss": 3.279735803604126,
"eval_runtime": 53.4471,
"eval_samples_per_second": 337.006,
"eval_steps_per_second": 21.068,
"step": 78000
},
{
"epoch": 21.00376830318691,
"grad_norm": 0.4437350332736969,
"learning_rate": 0.00028513935640231587,
"loss": 3.0092,
"step": 78050
},
{
"epoch": 21.017226528854437,
"grad_norm": 0.4182265102863312,
"learning_rate": 0.00028493739060185807,
"loss": 2.934,
"step": 78100
},
{
"epoch": 21.030684754521964,
"grad_norm": 0.43953824043273926,
"learning_rate": 0.00028473542480140026,
"loss": 2.947,
"step": 78150
},
{
"epoch": 21.04414298018949,
"grad_norm": 0.45151016116142273,
"learning_rate": 0.0002845334590009425,
"loss": 2.9428,
"step": 78200
},
{
"epoch": 21.05760120585702,
"grad_norm": 0.41173022985458374,
"learning_rate": 0.0002843314932004847,
"loss": 2.942,
"step": 78250
},
{
"epoch": 21.07105943152455,
"grad_norm": 0.4398365318775177,
"learning_rate": 0.0002841295274000269,
"loss": 2.9591,
"step": 78300
},
{
"epoch": 21.084517657192077,
"grad_norm": 0.43982040882110596,
"learning_rate": 0.0002839275615995691,
"loss": 2.9536,
"step": 78350
},
{
"epoch": 21.097975882859604,
"grad_norm": 0.43614324927330017,
"learning_rate": 0.0002837255957991113,
"loss": 2.9373,
"step": 78400
},
{
"epoch": 21.11143410852713,
"grad_norm": 0.41462191939353943,
"learning_rate": 0.00028352362999865355,
"loss": 2.9595,
"step": 78450
},
{
"epoch": 21.12489233419466,
"grad_norm": 0.4198111593723297,
"learning_rate": 0.00028332166419819574,
"loss": 2.9739,
"step": 78500
},
{
"epoch": 21.13835055986219,
"grad_norm": 0.4044225215911865,
"learning_rate": 0.000283119698397738,
"loss": 2.958,
"step": 78550
},
{
"epoch": 21.151808785529717,
"grad_norm": 0.4500187635421753,
"learning_rate": 0.0002829177325972802,
"loss": 2.9622,
"step": 78600
},
{
"epoch": 21.165267011197244,
"grad_norm": 0.4328445494174957,
"learning_rate": 0.0002827157667968224,
"loss": 2.9739,
"step": 78650
},
{
"epoch": 21.17872523686477,
"grad_norm": 0.4654206335544586,
"learning_rate": 0.0002825138009963646,
"loss": 2.9749,
"step": 78700
},
{
"epoch": 21.1921834625323,
"grad_norm": 0.4135824739933014,
"learning_rate": 0.0002823118351959068,
"loss": 2.9695,
"step": 78750
},
{
"epoch": 21.205641688199826,
"grad_norm": 0.4080086052417755,
"learning_rate": 0.00028210986939544903,
"loss": 2.9842,
"step": 78800
},
{
"epoch": 21.219099913867357,
"grad_norm": 0.421791672706604,
"learning_rate": 0.0002819079035949912,
"loss": 2.9721,
"step": 78850
},
{
"epoch": 21.232558139534884,
"grad_norm": 0.4322538673877716,
"learning_rate": 0.0002817059377945334,
"loss": 2.9641,
"step": 78900
},
{
"epoch": 21.24601636520241,
"grad_norm": 0.4447803795337677,
"learning_rate": 0.00028150397199407567,
"loss": 2.9708,
"step": 78950
},
{
"epoch": 21.25947459086994,
"grad_norm": 0.43925315141677856,
"learning_rate": 0.00028130200619361787,
"loss": 2.9669,
"step": 79000
},
{
"epoch": 21.25947459086994,
"eval_accuracy": 0.3955303987383058,
"eval_loss": 3.2910048961639404,
"eval_runtime": 53.1984,
"eval_samples_per_second": 338.582,
"eval_steps_per_second": 21.166,
"step": 79000
},
{
"epoch": 21.272932816537466,
"grad_norm": 0.4245566129684448,
"learning_rate": 0.00028110004039316006,
"loss": 2.9844,
"step": 79050
},
{
"epoch": 21.286391042204997,
"grad_norm": 0.41280364990234375,
"learning_rate": 0.00028089807459270226,
"loss": 2.9868,
"step": 79100
},
{
"epoch": 21.299849267872524,
"grad_norm": 0.4568192958831787,
"learning_rate": 0.0002806961087922445,
"loss": 2.9726,
"step": 79150
},
{
"epoch": 21.31330749354005,
"grad_norm": 0.4463944137096405,
"learning_rate": 0.0002804941429917867,
"loss": 2.9838,
"step": 79200
},
{
"epoch": 21.32676571920758,
"grad_norm": 0.45832765102386475,
"learning_rate": 0.0002802921771913289,
"loss": 2.9887,
"step": 79250
},
{
"epoch": 21.340223944875106,
"grad_norm": 0.43500322103500366,
"learning_rate": 0.0002800902113908711,
"loss": 2.9965,
"step": 79300
},
{
"epoch": 21.353682170542637,
"grad_norm": 0.41847607493400574,
"learning_rate": 0.0002798882455904133,
"loss": 2.977,
"step": 79350
},
{
"epoch": 21.367140396210164,
"grad_norm": 0.4514514207839966,
"learning_rate": 0.00027968627978995555,
"loss": 2.9806,
"step": 79400
},
{
"epoch": 21.38059862187769,
"grad_norm": 0.447410523891449,
"learning_rate": 0.00027948431398949774,
"loss": 2.9908,
"step": 79450
},
{
"epoch": 21.39405684754522,
"grad_norm": 0.4610724449157715,
"learning_rate": 0.00027928234818904,
"loss": 2.9668,
"step": 79500
},
{
"epoch": 21.407515073212746,
"grad_norm": 0.4488871693611145,
"learning_rate": 0.0002790803823885822,
"loss": 2.9921,
"step": 79550
},
{
"epoch": 21.420973298880277,
"grad_norm": 0.42067262530326843,
"learning_rate": 0.0002788784165881244,
"loss": 2.9879,
"step": 79600
},
{
"epoch": 21.434431524547804,
"grad_norm": 0.4374084770679474,
"learning_rate": 0.0002786764507876666,
"loss": 2.9853,
"step": 79650
},
{
"epoch": 21.44788975021533,
"grad_norm": 0.4477153718471527,
"learning_rate": 0.0002784744849872088,
"loss": 3.0006,
"step": 79700
},
{
"epoch": 21.46134797588286,
"grad_norm": 0.4401930868625641,
"learning_rate": 0.000278272519186751,
"loss": 2.995,
"step": 79750
},
{
"epoch": 21.474806201550386,
"grad_norm": 0.4580068290233612,
"learning_rate": 0.0002780705533862932,
"loss": 2.9824,
"step": 79800
},
{
"epoch": 21.488264427217917,
"grad_norm": 0.45493486523628235,
"learning_rate": 0.0002778685875858355,
"loss": 2.9976,
"step": 79850
},
{
"epoch": 21.501722652885444,
"grad_norm": 0.44258391857147217,
"learning_rate": 0.00027766662178537767,
"loss": 2.9964,
"step": 79900
},
{
"epoch": 21.51518087855297,
"grad_norm": 0.4550727307796478,
"learning_rate": 0.00027746465598491987,
"loss": 2.9849,
"step": 79950
},
{
"epoch": 21.5286391042205,
"grad_norm": 0.4246881902217865,
"learning_rate": 0.00027726269018446206,
"loss": 3.0057,
"step": 80000
},
{
"epoch": 21.5286391042205,
"eval_accuracy": 0.39579940835268335,
"eval_loss": 3.2892096042633057,
"eval_runtime": 53.3098,
"eval_samples_per_second": 337.874,
"eval_steps_per_second": 21.122,
"step": 80000
},
{
"epoch": 21.542097329888026,
"grad_norm": 0.4142609238624573,
"learning_rate": 0.00027706072438400426,
"loss": 2.9998,
"step": 80050
},
{
"epoch": 21.555555555555557,
"grad_norm": 0.43777963519096375,
"learning_rate": 0.0002768587585835465,
"loss": 2.997,
"step": 80100
},
{
"epoch": 21.569013781223084,
"grad_norm": 0.4332168996334076,
"learning_rate": 0.0002766567927830887,
"loss": 3.0043,
"step": 80150
},
{
"epoch": 21.58247200689061,
"grad_norm": 0.42009490728378296,
"learning_rate": 0.0002764548269826309,
"loss": 3.0102,
"step": 80200
},
{
"epoch": 21.59593023255814,
"grad_norm": 0.44434231519699097,
"learning_rate": 0.0002762528611821731,
"loss": 2.9915,
"step": 80250
},
{
"epoch": 21.609388458225666,
"grad_norm": 0.44381824135780334,
"learning_rate": 0.00027605089538171535,
"loss": 2.995,
"step": 80300
},
{
"epoch": 21.622846683893197,
"grad_norm": 0.42324790358543396,
"learning_rate": 0.00027584892958125754,
"loss": 3.0152,
"step": 80350
},
{
"epoch": 21.636304909560724,
"grad_norm": 0.44452613592147827,
"learning_rate": 0.0002756469637807998,
"loss": 3.013,
"step": 80400
},
{
"epoch": 21.649763135228252,
"grad_norm": 0.45783066749572754,
"learning_rate": 0.000275444997980342,
"loss": 3.0144,
"step": 80450
},
{
"epoch": 21.66322136089578,
"grad_norm": 0.4607763886451721,
"learning_rate": 0.0002752430321798842,
"loss": 3.0039,
"step": 80500
},
{
"epoch": 21.676679586563306,
"grad_norm": 0.42622271180152893,
"learning_rate": 0.0002750410663794264,
"loss": 3.0025,
"step": 80550
},
{
"epoch": 21.690137812230837,
"grad_norm": 0.4265972971916199,
"learning_rate": 0.0002748391005789686,
"loss": 3.0159,
"step": 80600
},
{
"epoch": 21.703596037898365,
"grad_norm": 0.4468063414096832,
"learning_rate": 0.00027463713477851083,
"loss": 3.007,
"step": 80650
},
{
"epoch": 21.717054263565892,
"grad_norm": 0.40498435497283936,
"learning_rate": 0.000274435168978053,
"loss": 3.0038,
"step": 80700
},
{
"epoch": 21.73051248923342,
"grad_norm": 0.48580971360206604,
"learning_rate": 0.0002742332031775953,
"loss": 3.0107,
"step": 80750
},
{
"epoch": 21.743970714900946,
"grad_norm": 0.4329358637332916,
"learning_rate": 0.00027403123737713747,
"loss": 3.0153,
"step": 80800
},
{
"epoch": 21.757428940568474,
"grad_norm": 0.4601568877696991,
"learning_rate": 0.00027382927157667967,
"loss": 3.0082,
"step": 80850
},
{
"epoch": 21.770887166236005,
"grad_norm": 0.4607238173484802,
"learning_rate": 0.00027362730577622186,
"loss": 3.0079,
"step": 80900
},
{
"epoch": 21.784345391903532,
"grad_norm": 0.44796958565711975,
"learning_rate": 0.00027342533997576406,
"loss": 3.0135,
"step": 80950
},
{
"epoch": 21.79780361757106,
"grad_norm": 0.47503456473350525,
"learning_rate": 0.0002732233741753063,
"loss": 3.0159,
"step": 81000
},
{
"epoch": 21.79780361757106,
"eval_accuracy": 0.3963195008502703,
"eval_loss": 3.2789855003356934,
"eval_runtime": 53.1425,
"eval_samples_per_second": 338.938,
"eval_steps_per_second": 21.188,
"step": 81000
},
{
"epoch": 21.811261843238587,
"grad_norm": 0.4426865875720978,
"learning_rate": 0.0002730214083748485,
"loss": 3.0087,
"step": 81050
},
{
"epoch": 21.824720068906114,
"grad_norm": 0.4314759373664856,
"learning_rate": 0.0002728194425743907,
"loss": 3.0165,
"step": 81100
},
{
"epoch": 21.838178294573645,
"grad_norm": 0.44944486021995544,
"learning_rate": 0.0002726174767739329,
"loss": 3.019,
"step": 81150
},
{
"epoch": 21.851636520241172,
"grad_norm": 0.4344472289085388,
"learning_rate": 0.00027241551097347515,
"loss": 3.011,
"step": 81200
},
{
"epoch": 21.8650947459087,
"grad_norm": 0.4304898679256439,
"learning_rate": 0.00027221354517301734,
"loss": 3.0056,
"step": 81250
},
{
"epoch": 21.878552971576227,
"grad_norm": 0.43768954277038574,
"learning_rate": 0.00027201157937255954,
"loss": 3.0129,
"step": 81300
},
{
"epoch": 21.892011197243754,
"grad_norm": 0.4472970962524414,
"learning_rate": 0.0002718096135721018,
"loss": 3.0209,
"step": 81350
},
{
"epoch": 21.905469422911285,
"grad_norm": 0.4112667739391327,
"learning_rate": 0.000271607647771644,
"loss": 3.015,
"step": 81400
},
{
"epoch": 21.918927648578812,
"grad_norm": 0.44268813729286194,
"learning_rate": 0.0002714056819711862,
"loss": 3.0118,
"step": 81450
},
{
"epoch": 21.93238587424634,
"grad_norm": 0.4346647560596466,
"learning_rate": 0.0002712037161707284,
"loss": 3.0199,
"step": 81500
},
{
"epoch": 21.945844099913867,
"grad_norm": 0.4291246831417084,
"learning_rate": 0.0002710017503702706,
"loss": 3.0145,
"step": 81550
},
{
"epoch": 21.959302325581394,
"grad_norm": 0.44748732447624207,
"learning_rate": 0.0002707997845698128,
"loss": 3.0253,
"step": 81600
},
{
"epoch": 21.972760551248925,
"grad_norm": 0.46326854825019836,
"learning_rate": 0.000270597818769355,
"loss": 3.0084,
"step": 81650
},
{
"epoch": 21.986218776916452,
"grad_norm": 0.4273764193058014,
"learning_rate": 0.00027039585296889727,
"loss": 3.0134,
"step": 81700
},
{
"epoch": 21.99967700258398,
"grad_norm": 0.4094850718975067,
"learning_rate": 0.00027019388716843947,
"loss": 3.0211,
"step": 81750
},
{
"epoch": 22.012919896640827,
"grad_norm": 0.4577467143535614,
"learning_rate": 0.00026999192136798166,
"loss": 2.9251,
"step": 81800
},
{
"epoch": 22.026378122308355,
"grad_norm": 0.4286164343357086,
"learning_rate": 0.00026978995556752386,
"loss": 2.9454,
"step": 81850
},
{
"epoch": 22.039836347975882,
"grad_norm": 0.46044474840164185,
"learning_rate": 0.00026958798976706606,
"loss": 2.9283,
"step": 81900
},
{
"epoch": 22.05329457364341,
"grad_norm": 0.463361918926239,
"learning_rate": 0.0002693860239666083,
"loss": 2.9374,
"step": 81950
},
{
"epoch": 22.06675279931094,
"grad_norm": 0.4062318503856659,
"learning_rate": 0.0002691840581661505,
"loss": 2.9459,
"step": 82000
},
{
"epoch": 22.06675279931094,
"eval_accuracy": 0.3958484080845429,
"eval_loss": 3.290665864944458,
"eval_runtime": 53.1747,
"eval_samples_per_second": 338.733,
"eval_steps_per_second": 21.175,
"step": 82000
},
{
"epoch": 22.080211024978468,
"grad_norm": 0.44609442353248596,
"learning_rate": 0.0002689820923656927,
"loss": 2.9442,
"step": 82050
},
{
"epoch": 22.093669250645995,
"grad_norm": 0.4272725582122803,
"learning_rate": 0.00026878012656523495,
"loss": 2.9463,
"step": 82100
},
{
"epoch": 22.107127476313522,
"grad_norm": 0.47067728638648987,
"learning_rate": 0.00026857816076477714,
"loss": 2.9489,
"step": 82150
},
{
"epoch": 22.12058570198105,
"grad_norm": 0.46830788254737854,
"learning_rate": 0.00026837619496431934,
"loss": 2.9499,
"step": 82200
},
{
"epoch": 22.13404392764858,
"grad_norm": 0.41217130422592163,
"learning_rate": 0.0002681742291638616,
"loss": 2.9487,
"step": 82250
},
{
"epoch": 22.147502153316108,
"grad_norm": 0.4633793830871582,
"learning_rate": 0.0002679722633634038,
"loss": 2.9559,
"step": 82300
},
{
"epoch": 22.160960378983635,
"grad_norm": 0.4432108402252197,
"learning_rate": 0.000267770297562946,
"loss": 2.9572,
"step": 82350
},
{
"epoch": 22.174418604651162,
"grad_norm": 0.4489414393901825,
"learning_rate": 0.0002675683317624882,
"loss": 2.9574,
"step": 82400
},
{
"epoch": 22.18787683031869,
"grad_norm": 0.4502296447753906,
"learning_rate": 0.0002673663659620304,
"loss": 2.9509,
"step": 82450
},
{
"epoch": 22.20133505598622,
"grad_norm": 0.43019405007362366,
"learning_rate": 0.0002671644001615726,
"loss": 2.9554,
"step": 82500
},
{
"epoch": 22.214793281653748,
"grad_norm": 0.44396713376045227,
"learning_rate": 0.0002669624343611148,
"loss": 2.9621,
"step": 82550
},
{
"epoch": 22.228251507321275,
"grad_norm": 0.47302505373954773,
"learning_rate": 0.00026676046856065707,
"loss": 2.9714,
"step": 82600
},
{
"epoch": 22.241709732988802,
"grad_norm": 0.4609719216823578,
"learning_rate": 0.00026655850276019927,
"loss": 2.9692,
"step": 82650
},
{
"epoch": 22.25516795865633,
"grad_norm": 0.45992007851600647,
"learning_rate": 0.00026635653695974146,
"loss": 2.9599,
"step": 82700
},
{
"epoch": 22.26862618432386,
"grad_norm": 0.4323056936264038,
"learning_rate": 0.00026615457115928366,
"loss": 2.9689,
"step": 82750
},
{
"epoch": 22.282084409991388,
"grad_norm": 0.4838802218437195,
"learning_rate": 0.00026595260535882586,
"loss": 2.9716,
"step": 82800
},
{
"epoch": 22.295542635658915,
"grad_norm": 0.44049209356307983,
"learning_rate": 0.0002657506395583681,
"loss": 2.9746,
"step": 82850
},
{
"epoch": 22.309000861326442,
"grad_norm": 0.44948625564575195,
"learning_rate": 0.0002655486737579103,
"loss": 2.9831,
"step": 82900
},
{
"epoch": 22.32245908699397,
"grad_norm": 0.4355025291442871,
"learning_rate": 0.00026534670795745255,
"loss": 2.9641,
"step": 82950
},
{
"epoch": 22.3359173126615,
"grad_norm": 0.44514182209968567,
"learning_rate": 0.00026514474215699475,
"loss": 2.9807,
"step": 83000
},
{
"epoch": 22.3359173126615,
"eval_accuracy": 0.3954855275869577,
"eval_loss": 3.2914698123931885,
"eval_runtime": 53.3839,
"eval_samples_per_second": 337.405,
"eval_steps_per_second": 21.092,
"step": 83000
},
{
"epoch": 22.349375538329028,
"grad_norm": 0.4393146336078644,
"learning_rate": 0.00026494277635653694,
"loss": 2.9734,
"step": 83050
},
{
"epoch": 22.362833763996555,
"grad_norm": 0.44433945417404175,
"learning_rate": 0.00026474081055607914,
"loss": 2.9804,
"step": 83100
},
{
"epoch": 22.376291989664082,
"grad_norm": 0.4565770626068115,
"learning_rate": 0.00026453884475562134,
"loss": 2.9793,
"step": 83150
},
{
"epoch": 22.38975021533161,
"grad_norm": 0.45272794365882874,
"learning_rate": 0.0002643368789551636,
"loss": 2.9851,
"step": 83200
},
{
"epoch": 22.403208440999137,
"grad_norm": 0.47611159086227417,
"learning_rate": 0.0002641349131547058,
"loss": 2.9829,
"step": 83250
},
{
"epoch": 22.416666666666668,
"grad_norm": 0.43398815393447876,
"learning_rate": 0.000263932947354248,
"loss": 2.9736,
"step": 83300
},
{
"epoch": 22.430124892334195,
"grad_norm": 0.4315131902694702,
"learning_rate": 0.0002637309815537902,
"loss": 2.9792,
"step": 83350
},
{
"epoch": 22.443583118001722,
"grad_norm": 0.4292556941509247,
"learning_rate": 0.0002635290157533324,
"loss": 2.9753,
"step": 83400
},
{
"epoch": 22.45704134366925,
"grad_norm": 0.42346417903900146,
"learning_rate": 0.0002633270499528746,
"loss": 2.9812,
"step": 83450
},
{
"epoch": 22.470499569336777,
"grad_norm": 0.4710681438446045,
"learning_rate": 0.0002631250841524168,
"loss": 2.9835,
"step": 83500
},
{
"epoch": 22.483957795004308,
"grad_norm": 0.42093920707702637,
"learning_rate": 0.00026292311835195907,
"loss": 2.9884,
"step": 83550
},
{
"epoch": 22.497416020671835,
"grad_norm": 0.4511464238166809,
"learning_rate": 0.00026272115255150126,
"loss": 2.9923,
"step": 83600
},
{
"epoch": 22.510874246339363,
"grad_norm": 0.4473262429237366,
"learning_rate": 0.00026251918675104346,
"loss": 2.9751,
"step": 83650
},
{
"epoch": 22.52433247200689,
"grad_norm": 0.41240695118904114,
"learning_rate": 0.00026231722095058566,
"loss": 2.9958,
"step": 83700
},
{
"epoch": 22.537790697674417,
"grad_norm": 0.4413813352584839,
"learning_rate": 0.00026211525515012785,
"loss": 2.979,
"step": 83750
},
{
"epoch": 22.551248923341948,
"grad_norm": 0.4364555776119232,
"learning_rate": 0.0002619132893496701,
"loss": 2.9978,
"step": 83800
},
{
"epoch": 22.564707149009475,
"grad_norm": 0.44257232546806335,
"learning_rate": 0.0002617113235492123,
"loss": 2.9917,
"step": 83850
},
{
"epoch": 22.578165374677003,
"grad_norm": 0.4314466416835785,
"learning_rate": 0.00026150935774875455,
"loss": 2.9873,
"step": 83900
},
{
"epoch": 22.59162360034453,
"grad_norm": 0.4465186297893524,
"learning_rate": 0.00026130739194829675,
"loss": 2.9795,
"step": 83950
},
{
"epoch": 22.605081826012057,
"grad_norm": 0.46384963393211365,
"learning_rate": 0.00026110542614783894,
"loss": 2.9952,
"step": 84000
},
{
"epoch": 22.605081826012057,
"eval_accuracy": 0.3961026417265637,
"eval_loss": 3.283360242843628,
"eval_runtime": 53.2066,
"eval_samples_per_second": 338.53,
"eval_steps_per_second": 21.163,
"step": 84000
},
{
"epoch": 22.618540051679588,
"grad_norm": 0.4887400269508362,
"learning_rate": 0.00026090346034738114,
"loss": 2.9848,
"step": 84050
},
{
"epoch": 22.631998277347115,
"grad_norm": 0.5525014400482178,
"learning_rate": 0.0002607014945469234,
"loss": 2.9975,
"step": 84100
},
{
"epoch": 22.645456503014643,
"grad_norm": 0.4679363965988159,
"learning_rate": 0.0002604995287464656,
"loss": 2.9913,
"step": 84150
},
{
"epoch": 22.65891472868217,
"grad_norm": 0.43187206983566284,
"learning_rate": 0.0002602975629460078,
"loss": 2.995,
"step": 84200
},
{
"epoch": 22.672372954349697,
"grad_norm": 0.4239598512649536,
"learning_rate": 0.00026009559714555,
"loss": 2.9935,
"step": 84250
},
{
"epoch": 22.685831180017228,
"grad_norm": 0.4313739240169525,
"learning_rate": 0.0002598936313450922,
"loss": 2.9888,
"step": 84300
},
{
"epoch": 22.699289405684755,
"grad_norm": 0.45418158173561096,
"learning_rate": 0.0002596916655446344,
"loss": 2.9967,
"step": 84350
},
{
"epoch": 22.712747631352283,
"grad_norm": 0.4296848475933075,
"learning_rate": 0.0002594896997441766,
"loss": 2.9942,
"step": 84400
},
{
"epoch": 22.72620585701981,
"grad_norm": 0.4835188686847687,
"learning_rate": 0.00025928773394371887,
"loss": 2.9931,
"step": 84450
},
{
"epoch": 22.739664082687337,
"grad_norm": 0.44101449847221375,
"learning_rate": 0.00025908576814326107,
"loss": 3.0047,
"step": 84500
},
{
"epoch": 22.753122308354868,
"grad_norm": 0.42555737495422363,
"learning_rate": 0.00025888380234280326,
"loss": 2.9982,
"step": 84550
},
{
"epoch": 22.766580534022395,
"grad_norm": 0.4653916358947754,
"learning_rate": 0.00025868183654234546,
"loss": 2.9946,
"step": 84600
},
{
"epoch": 22.780038759689923,
"grad_norm": 0.4454815089702606,
"learning_rate": 0.00025847987074188765,
"loss": 3.0072,
"step": 84650
},
{
"epoch": 22.79349698535745,
"grad_norm": 0.45986753702163696,
"learning_rate": 0.0002582779049414299,
"loss": 3.0039,
"step": 84700
},
{
"epoch": 22.806955211024977,
"grad_norm": 0.4703720510005951,
"learning_rate": 0.0002580759391409721,
"loss": 2.9963,
"step": 84750
},
{
"epoch": 22.820413436692505,
"grad_norm": 0.45785728096961975,
"learning_rate": 0.00025787397334051435,
"loss": 3.001,
"step": 84800
},
{
"epoch": 22.833871662360036,
"grad_norm": 0.454120010137558,
"learning_rate": 0.00025767200754005655,
"loss": 2.994,
"step": 84850
},
{
"epoch": 22.847329888027563,
"grad_norm": 0.4418194890022278,
"learning_rate": 0.00025747004173959874,
"loss": 3.0124,
"step": 84900
},
{
"epoch": 22.86078811369509,
"grad_norm": 0.4594516456127167,
"learning_rate": 0.00025726807593914094,
"loss": 3.0028,
"step": 84950
},
{
"epoch": 22.874246339362617,
"grad_norm": 0.42217785120010376,
"learning_rate": 0.00025706611013868313,
"loss": 2.9981,
"step": 85000
},
{
"epoch": 22.874246339362617,
"eval_accuracy": 0.39679091955656437,
"eval_loss": 3.2778079509735107,
"eval_runtime": 53.2599,
"eval_samples_per_second": 338.19,
"eval_steps_per_second": 21.142,
"step": 85000
},
{
"epoch": 22.887704565030145,
"grad_norm": 0.43962785601615906,
"learning_rate": 0.0002568641443382254,
"loss": 3.0046,
"step": 85050
},
{
"epoch": 22.901162790697676,
"grad_norm": 0.4394022226333618,
"learning_rate": 0.0002566621785377676,
"loss": 2.999,
"step": 85100
},
{
"epoch": 22.914621016365203,
"grad_norm": 0.441394567489624,
"learning_rate": 0.0002564602127373098,
"loss": 3.0038,
"step": 85150
},
{
"epoch": 22.92807924203273,
"grad_norm": 0.4967244863510132,
"learning_rate": 0.00025625824693685203,
"loss": 3.0086,
"step": 85200
},
{
"epoch": 22.941537467700257,
"grad_norm": 0.44626763463020325,
"learning_rate": 0.0002560562811363942,
"loss": 3.014,
"step": 85250
},
{
"epoch": 22.954995693367785,
"grad_norm": 0.43168094754219055,
"learning_rate": 0.0002558543153359364,
"loss": 2.9938,
"step": 85300
},
{
"epoch": 22.968453919035316,
"grad_norm": 0.45025435090065,
"learning_rate": 0.0002556523495354786,
"loss": 3.013,
"step": 85350
},
{
"epoch": 22.981912144702843,
"grad_norm": 0.42971986532211304,
"learning_rate": 0.00025545038373502087,
"loss": 3.0124,
"step": 85400
},
{
"epoch": 22.99537037037037,
"grad_norm": 0.4267427921295166,
"learning_rate": 0.00025524841793456306,
"loss": 3.01,
"step": 85450
},
{
"epoch": 23.00861326442722,
"grad_norm": 0.4329337477684021,
"learning_rate": 0.00025504645213410526,
"loss": 2.9676,
"step": 85500
},
{
"epoch": 23.022071490094746,
"grad_norm": 0.4235066771507263,
"learning_rate": 0.00025484448633364745,
"loss": 2.9191,
"step": 85550
},
{
"epoch": 23.035529715762273,
"grad_norm": 0.47007474303245544,
"learning_rate": 0.00025464252053318965,
"loss": 2.9197,
"step": 85600
},
{
"epoch": 23.0489879414298,
"grad_norm": 0.462991327047348,
"learning_rate": 0.0002544405547327319,
"loss": 2.935,
"step": 85650
},
{
"epoch": 23.06244616709733,
"grad_norm": 0.4521004855632782,
"learning_rate": 0.0002542385889322741,
"loss": 2.922,
"step": 85700
},
{
"epoch": 23.07590439276486,
"grad_norm": 0.4741455614566803,
"learning_rate": 0.00025403662313181635,
"loss": 2.9292,
"step": 85750
},
{
"epoch": 23.089362618432386,
"grad_norm": 0.4532252550125122,
"learning_rate": 0.00025383465733135854,
"loss": 2.9462,
"step": 85800
},
{
"epoch": 23.102820844099913,
"grad_norm": 0.4494078457355499,
"learning_rate": 0.00025363269153090074,
"loss": 2.9332,
"step": 85850
},
{
"epoch": 23.11627906976744,
"grad_norm": 0.45025596022605896,
"learning_rate": 0.00025343072573044294,
"loss": 2.9574,
"step": 85900
},
{
"epoch": 23.12973729543497,
"grad_norm": 0.44868040084838867,
"learning_rate": 0.00025322875992998513,
"loss": 2.9478,
"step": 85950
},
{
"epoch": 23.1431955211025,
"grad_norm": 0.4450909495353699,
"learning_rate": 0.0002530267941295274,
"loss": 2.938,
"step": 86000
},
{
"epoch": 23.1431955211025,
"eval_accuracy": 0.3958298294722414,
"eval_loss": 3.2935590744018555,
"eval_runtime": 53.3991,
"eval_samples_per_second": 337.309,
"eval_steps_per_second": 21.087,
"step": 86000
},
{
"epoch": 23.156653746770026,
"grad_norm": 0.47848260402679443,
"learning_rate": 0.0002528248283290696,
"loss": 2.9541,
"step": 86050
},
{
"epoch": 23.170111972437553,
"grad_norm": 0.44607311487197876,
"learning_rate": 0.00025262286252861183,
"loss": 2.946,
"step": 86100
},
{
"epoch": 23.18357019810508,
"grad_norm": 0.44378456473350525,
"learning_rate": 0.000252420896728154,
"loss": 2.939,
"step": 86150
},
{
"epoch": 23.19702842377261,
"grad_norm": 0.4426654577255249,
"learning_rate": 0.0002522189309276962,
"loss": 2.9498,
"step": 86200
},
{
"epoch": 23.21048664944014,
"grad_norm": 0.45932042598724365,
"learning_rate": 0.0002520169651272384,
"loss": 2.9454,
"step": 86250
},
{
"epoch": 23.223944875107666,
"grad_norm": 0.45173564553260803,
"learning_rate": 0.00025181499932678067,
"loss": 2.957,
"step": 86300
},
{
"epoch": 23.237403100775193,
"grad_norm": 0.44965678453445435,
"learning_rate": 0.00025161303352632286,
"loss": 2.952,
"step": 86350
},
{
"epoch": 23.25086132644272,
"grad_norm": 0.4480496048927307,
"learning_rate": 0.00025141106772586506,
"loss": 2.9572,
"step": 86400
},
{
"epoch": 23.26431955211025,
"grad_norm": 0.46193182468414307,
"learning_rate": 0.00025120910192540726,
"loss": 2.9526,
"step": 86450
},
{
"epoch": 23.27777777777778,
"grad_norm": 0.48119649291038513,
"learning_rate": 0.00025100713612494945,
"loss": 2.9621,
"step": 86500
},
{
"epoch": 23.291236003445306,
"grad_norm": 0.4442594647407532,
"learning_rate": 0.0002508051703244917,
"loss": 2.9495,
"step": 86550
},
{
"epoch": 23.304694229112833,
"grad_norm": 0.4543590247631073,
"learning_rate": 0.0002506032045240339,
"loss": 2.9681,
"step": 86600
},
{
"epoch": 23.31815245478036,
"grad_norm": 0.4707615375518799,
"learning_rate": 0.00025040123872357615,
"loss": 2.9575,
"step": 86650
},
{
"epoch": 23.33161068044789,
"grad_norm": 0.4720620810985565,
"learning_rate": 0.00025019927292311834,
"loss": 2.9659,
"step": 86700
},
{
"epoch": 23.34506890611542,
"grad_norm": 0.47484633326530457,
"learning_rate": 0.00024999730712266054,
"loss": 2.9692,
"step": 86750
},
{
"epoch": 23.358527131782946,
"grad_norm": 0.44391027092933655,
"learning_rate": 0.00024979534132220274,
"loss": 2.9715,
"step": 86800
},
{
"epoch": 23.371985357450473,
"grad_norm": 0.4550620913505554,
"learning_rate": 0.00024959337552174493,
"loss": 2.974,
"step": 86850
},
{
"epoch": 23.385443583118,
"grad_norm": 0.41127505898475647,
"learning_rate": 0.0002493914097212872,
"loss": 2.9671,
"step": 86900
},
{
"epoch": 23.39890180878553,
"grad_norm": 0.4839131236076355,
"learning_rate": 0.0002491894439208294,
"loss": 2.968,
"step": 86950
},
{
"epoch": 23.41236003445306,
"grad_norm": 0.4719763696193695,
"learning_rate": 0.00024898747812037163,
"loss": 2.9782,
"step": 87000
},
{
"epoch": 23.41236003445306,
"eval_accuracy": 0.3960650499145384,
"eval_loss": 3.2877259254455566,
"eval_runtime": 53.193,
"eval_samples_per_second": 338.616,
"eval_steps_per_second": 21.168,
"step": 87000
},
{
"epoch": 23.425818260120586,
"grad_norm": 0.4370049238204956,
"learning_rate": 0.0002487855123199138,
"loss": 2.9755,
"step": 87050
},
{
"epoch": 23.439276485788113,
"grad_norm": 0.424113005399704,
"learning_rate": 0.000248583546519456,
"loss": 2.9699,
"step": 87100
},
{
"epoch": 23.45273471145564,
"grad_norm": 0.4651334285736084,
"learning_rate": 0.0002483815807189982,
"loss": 2.9692,
"step": 87150
},
{
"epoch": 23.466192937123168,
"grad_norm": 0.4388973116874695,
"learning_rate": 0.0002481796149185404,
"loss": 2.9781,
"step": 87200
},
{
"epoch": 23.4796511627907,
"grad_norm": 0.4796842634677887,
"learning_rate": 0.00024797764911808266,
"loss": 2.9739,
"step": 87250
},
{
"epoch": 23.493109388458226,
"grad_norm": 0.4598848819732666,
"learning_rate": 0.00024777568331762486,
"loss": 2.9593,
"step": 87300
},
{
"epoch": 23.506567614125753,
"grad_norm": 0.4424052834510803,
"learning_rate": 0.00024757371751716706,
"loss": 2.9831,
"step": 87350
},
{
"epoch": 23.52002583979328,
"grad_norm": 0.43741047382354736,
"learning_rate": 0.0002473717517167093,
"loss": 2.9668,
"step": 87400
},
{
"epoch": 23.533484065460808,
"grad_norm": 0.4539873003959656,
"learning_rate": 0.0002471697859162515,
"loss": 2.9709,
"step": 87450
},
{
"epoch": 23.54694229112834,
"grad_norm": 0.4834721088409424,
"learning_rate": 0.0002469678201157937,
"loss": 2.9651,
"step": 87500
},
{
"epoch": 23.560400516795866,
"grad_norm": 0.4531641900539398,
"learning_rate": 0.0002467658543153359,
"loss": 2.9698,
"step": 87550
},
{
"epoch": 23.573858742463393,
"grad_norm": 0.4506773054599762,
"learning_rate": 0.00024656388851487815,
"loss": 2.976,
"step": 87600
},
{
"epoch": 23.58731696813092,
"grad_norm": 0.49148374795913696,
"learning_rate": 0.00024636192271442034,
"loss": 2.9882,
"step": 87650
},
{
"epoch": 23.600775193798448,
"grad_norm": 0.48155003786087036,
"learning_rate": 0.00024615995691396254,
"loss": 2.992,
"step": 87700
},
{
"epoch": 23.61423341946598,
"grad_norm": 0.4349481165409088,
"learning_rate": 0.00024595799111350473,
"loss": 2.9826,
"step": 87750
},
{
"epoch": 23.627691645133506,
"grad_norm": 0.4904167652130127,
"learning_rate": 0.00024575602531304693,
"loss": 2.9799,
"step": 87800
},
{
"epoch": 23.641149870801033,
"grad_norm": 0.44706079363822937,
"learning_rate": 0.0002455540595125892,
"loss": 2.9929,
"step": 87850
},
{
"epoch": 23.65460809646856,
"grad_norm": 0.458001047372818,
"learning_rate": 0.00024535209371213143,
"loss": 2.9725,
"step": 87900
},
{
"epoch": 23.668066322136088,
"grad_norm": 0.4497480094432831,
"learning_rate": 0.0002451501279116736,
"loss": 2.9881,
"step": 87950
},
{
"epoch": 23.68152454780362,
"grad_norm": 0.4559875726699829,
"learning_rate": 0.0002449481621112158,
"loss": 2.9777,
"step": 88000
},
{
"epoch": 23.68152454780362,
"eval_accuracy": 0.39634253398364994,
"eval_loss": 3.282762050628662,
"eval_runtime": 53.172,
"eval_samples_per_second": 338.75,
"eval_steps_per_second": 21.177,
"step": 88000
},
{
"epoch": 23.694982773471146,
"grad_norm": 0.4412843883037567,
"learning_rate": 0.000244746196310758,
"loss": 2.9958,
"step": 88050
},
{
"epoch": 23.708440999138674,
"grad_norm": 0.4425804913043976,
"learning_rate": 0.0002445442305103002,
"loss": 2.9793,
"step": 88100
},
{
"epoch": 23.7218992248062,
"grad_norm": 0.4500211775302887,
"learning_rate": 0.00024434226470984246,
"loss": 2.9853,
"step": 88150
},
{
"epoch": 23.735357450473728,
"grad_norm": 0.4496285319328308,
"learning_rate": 0.00024414029890938463,
"loss": 2.9882,
"step": 88200
},
{
"epoch": 23.74881567614126,
"grad_norm": 0.47043538093566895,
"learning_rate": 0.00024393833310892686,
"loss": 2.9939,
"step": 88250
},
{
"epoch": 23.762273901808786,
"grad_norm": 0.4536699652671814,
"learning_rate": 0.0002437363673084691,
"loss": 2.984,
"step": 88300
},
{
"epoch": 23.775732127476314,
"grad_norm": 0.46850210428237915,
"learning_rate": 0.0002435344015080113,
"loss": 2.9818,
"step": 88350
},
{
"epoch": 23.78919035314384,
"grad_norm": 0.47080132365226746,
"learning_rate": 0.0002433324357075535,
"loss": 2.9908,
"step": 88400
},
{
"epoch": 23.802648578811368,
"grad_norm": 0.42550772428512573,
"learning_rate": 0.00024313046990709572,
"loss": 2.9861,
"step": 88450
},
{
"epoch": 23.8161068044789,
"grad_norm": 0.45124003291130066,
"learning_rate": 0.00024292850410663792,
"loss": 2.9984,
"step": 88500
},
{
"epoch": 23.829565030146426,
"grad_norm": 0.4718407690525055,
"learning_rate": 0.00024272653830618014,
"loss": 2.9919,
"step": 88550
},
{
"epoch": 23.843023255813954,
"grad_norm": 0.45113521814346313,
"learning_rate": 0.00024252457250572234,
"loss": 2.9983,
"step": 88600
},
{
"epoch": 23.85648148148148,
"grad_norm": 0.4578773081302643,
"learning_rate": 0.00024232260670526453,
"loss": 2.993,
"step": 88650
},
{
"epoch": 23.86993970714901,
"grad_norm": 0.43773236870765686,
"learning_rate": 0.00024212064090480676,
"loss": 2.9961,
"step": 88700
},
{
"epoch": 23.88339793281654,
"grad_norm": 0.46326008439064026,
"learning_rate": 0.00024191867510434898,
"loss": 3.0035,
"step": 88750
},
{
"epoch": 23.896856158484066,
"grad_norm": 0.45599448680877686,
"learning_rate": 0.0002417167093038912,
"loss": 2.9981,
"step": 88800
},
{
"epoch": 23.910314384151594,
"grad_norm": 0.49931249022483826,
"learning_rate": 0.0002415147435034334,
"loss": 2.9917,
"step": 88850
},
{
"epoch": 23.92377260981912,
"grad_norm": 0.45583376288414,
"learning_rate": 0.00024131277770297562,
"loss": 2.9875,
"step": 88900
},
{
"epoch": 23.93723083548665,
"grad_norm": 0.44669798016548157,
"learning_rate": 0.00024111081190251782,
"loss": 2.9896,
"step": 88950
},
{
"epoch": 23.95068906115418,
"grad_norm": 0.4412989616394043,
"learning_rate": 0.00024090884610206002,
"loss": 3.0013,
"step": 89000
},
{
"epoch": 23.95068906115418,
"eval_accuracy": 0.39690010964640665,
"eval_loss": 3.275252103805542,
"eval_runtime": 53.4152,
"eval_samples_per_second": 337.207,
"eval_steps_per_second": 21.08,
"step": 89000
},
{
"epoch": 23.964147286821706,
"grad_norm": 0.4542916715145111,
"learning_rate": 0.00024070688030160224,
"loss": 2.9937,
"step": 89050
},
{
"epoch": 23.977605512489234,
"grad_norm": 0.4350804090499878,
"learning_rate": 0.00024050491450114443,
"loss": 2.9948,
"step": 89100
},
{
"epoch": 23.99106373815676,
"grad_norm": 0.49811744689941406,
"learning_rate": 0.00024030294870068666,
"loss": 2.9902,
"step": 89150
},
{
"epoch": 24.00430663221361,
"grad_norm": 0.4629533290863037,
"learning_rate": 0.00024010098290022888,
"loss": 2.9727,
"step": 89200
},
{
"epoch": 24.017764857881136,
"grad_norm": 0.4568398594856262,
"learning_rate": 0.0002398990170997711,
"loss": 2.9066,
"step": 89250
},
{
"epoch": 24.031223083548664,
"grad_norm": 0.46679773926734924,
"learning_rate": 0.0002396970512993133,
"loss": 2.9074,
"step": 89300
},
{
"epoch": 24.044681309216195,
"grad_norm": 0.4443589746952057,
"learning_rate": 0.00023949508549885552,
"loss": 2.9111,
"step": 89350
},
{
"epoch": 24.058139534883722,
"grad_norm": 0.4438439607620239,
"learning_rate": 0.00023929311969839772,
"loss": 2.92,
"step": 89400
},
{
"epoch": 24.07159776055125,
"grad_norm": 0.43250682950019836,
"learning_rate": 0.00023909115389793992,
"loss": 2.9236,
"step": 89450
},
{
"epoch": 24.085055986218777,
"grad_norm": 0.45004555583000183,
"learning_rate": 0.00023888918809748214,
"loss": 2.931,
"step": 89500
},
{
"epoch": 24.098514211886304,
"grad_norm": 0.47898465394973755,
"learning_rate": 0.00023868722229702433,
"loss": 2.9268,
"step": 89550
},
{
"epoch": 24.11197243755383,
"grad_norm": 0.46386006474494934,
"learning_rate": 0.00023848525649656656,
"loss": 2.943,
"step": 89600
},
{
"epoch": 24.125430663221362,
"grad_norm": 0.43041688203811646,
"learning_rate": 0.00023828329069610878,
"loss": 2.9324,
"step": 89650
},
{
"epoch": 24.13888888888889,
"grad_norm": 0.45035499334335327,
"learning_rate": 0.000238081324895651,
"loss": 2.9312,
"step": 89700
},
{
"epoch": 24.152347114556417,
"grad_norm": 0.47834378480911255,
"learning_rate": 0.0002378793590951932,
"loss": 2.9443,
"step": 89750
},
{
"epoch": 24.165805340223944,
"grad_norm": 0.4648357033729553,
"learning_rate": 0.0002376773932947354,
"loss": 2.9397,
"step": 89800
},
{
"epoch": 24.17926356589147,
"grad_norm": 0.45432841777801514,
"learning_rate": 0.00023747542749427762,
"loss": 2.9383,
"step": 89850
},
{
"epoch": 24.192721791559002,
"grad_norm": 0.451199471950531,
"learning_rate": 0.00023727346169381982,
"loss": 2.9324,
"step": 89900
},
{
"epoch": 24.20618001722653,
"grad_norm": 0.4608922302722931,
"learning_rate": 0.00023707149589336204,
"loss": 2.9457,
"step": 89950
},
{
"epoch": 24.219638242894057,
"grad_norm": 0.44181180000305176,
"learning_rate": 0.00023686953009290424,
"loss": 2.9362,
"step": 90000
},
{
"epoch": 24.219638242894057,
"eval_accuracy": 0.3963043989373468,
"eval_loss": 3.2903923988342285,
"eval_runtime": 53.3682,
"eval_samples_per_second": 337.505,
"eval_steps_per_second": 21.099,
"step": 90000
},
{
"epoch": 24.233096468561584,
"grad_norm": 0.4413633942604065,
"learning_rate": 0.00023666756429244643,
"loss": 2.9332,
"step": 90050
},
{
"epoch": 24.24655469422911,
"grad_norm": 0.4630167484283447,
"learning_rate": 0.00023646559849198868,
"loss": 2.9405,
"step": 90100
},
{
"epoch": 24.260012919896642,
"grad_norm": 0.47358977794647217,
"learning_rate": 0.00023626363269153088,
"loss": 2.9412,
"step": 90150
},
{
"epoch": 24.27347114556417,
"grad_norm": 0.4817167818546295,
"learning_rate": 0.0002360616668910731,
"loss": 2.9531,
"step": 90200
},
{
"epoch": 24.286929371231697,
"grad_norm": 0.5337883830070496,
"learning_rate": 0.0002358597010906153,
"loss": 2.9484,
"step": 90250
},
{
"epoch": 24.300387596899224,
"grad_norm": 0.4577960968017578,
"learning_rate": 0.00023565773529015752,
"loss": 2.956,
"step": 90300
},
{
"epoch": 24.31384582256675,
"grad_norm": 0.45729923248291016,
"learning_rate": 0.00023545576948969972,
"loss": 2.9503,
"step": 90350
},
{
"epoch": 24.327304048234282,
"grad_norm": 0.4802705943584442,
"learning_rate": 0.00023525380368924194,
"loss": 2.9615,
"step": 90400
},
{
"epoch": 24.34076227390181,
"grad_norm": 0.4446525275707245,
"learning_rate": 0.00023505183788878414,
"loss": 2.9507,
"step": 90450
},
{
"epoch": 24.354220499569337,
"grad_norm": 0.437545508146286,
"learning_rate": 0.00023484987208832633,
"loss": 2.9511,
"step": 90500
},
{
"epoch": 24.367678725236864,
"grad_norm": 0.46033021807670593,
"learning_rate": 0.00023464790628786858,
"loss": 2.9563,
"step": 90550
},
{
"epoch": 24.38113695090439,
"grad_norm": 0.4750497341156006,
"learning_rate": 0.00023444594048741078,
"loss": 2.9652,
"step": 90600
},
{
"epoch": 24.394595176571922,
"grad_norm": 0.4922107458114624,
"learning_rate": 0.000234243974686953,
"loss": 2.9485,
"step": 90650
},
{
"epoch": 24.40805340223945,
"grad_norm": 0.4701991379261017,
"learning_rate": 0.0002340420088864952,
"loss": 2.9618,
"step": 90700
},
{
"epoch": 24.421511627906977,
"grad_norm": 0.46807754039764404,
"learning_rate": 0.00023384004308603742,
"loss": 2.982,
"step": 90750
},
{
"epoch": 24.434969853574504,
"grad_norm": 0.4745599627494812,
"learning_rate": 0.00023363807728557962,
"loss": 2.9547,
"step": 90800
},
{
"epoch": 24.44842807924203,
"grad_norm": 0.4703523516654968,
"learning_rate": 0.0002334361114851218,
"loss": 2.9533,
"step": 90850
},
{
"epoch": 24.461886304909562,
"grad_norm": 0.49658453464508057,
"learning_rate": 0.00023323414568466404,
"loss": 2.9641,
"step": 90900
},
{
"epoch": 24.47534453057709,
"grad_norm": 0.4539616107940674,
"learning_rate": 0.00023303217988420623,
"loss": 2.9629,
"step": 90950
},
{
"epoch": 24.488802756244617,
"grad_norm": 0.5139576196670532,
"learning_rate": 0.00023283021408374848,
"loss": 2.951,
"step": 91000
},
{
"epoch": 24.488802756244617,
"eval_accuracy": 0.3967428976464049,
"eval_loss": 3.28558087348938,
"eval_runtime": 53.262,
"eval_samples_per_second": 338.177,
"eval_steps_per_second": 21.141,
"step": 91000
},
{
"epoch": 24.502260981912144,
"grad_norm": 0.44592607021331787,
"learning_rate": 0.00023262824828329068,
"loss": 2.9816,
"step": 91050
},
{
"epoch": 24.51571920757967,
"grad_norm": 0.46577224135398865,
"learning_rate": 0.0002324262824828329,
"loss": 2.9624,
"step": 91100
},
{
"epoch": 24.529177433247202,
"grad_norm": 0.46494442224502563,
"learning_rate": 0.0002322243166823751,
"loss": 2.9691,
"step": 91150
},
{
"epoch": 24.54263565891473,
"grad_norm": 0.4683932662010193,
"learning_rate": 0.0002320223508819173,
"loss": 2.9675,
"step": 91200
},
{
"epoch": 24.556093884582257,
"grad_norm": 0.4711366593837738,
"learning_rate": 0.00023182038508145952,
"loss": 2.9824,
"step": 91250
},
{
"epoch": 24.569552110249784,
"grad_norm": 0.4714370369911194,
"learning_rate": 0.0002316184192810017,
"loss": 2.99,
"step": 91300
},
{
"epoch": 24.58301033591731,
"grad_norm": 0.4560222625732422,
"learning_rate": 0.00023141645348054394,
"loss": 2.9751,
"step": 91350
},
{
"epoch": 24.596468561584842,
"grad_norm": 0.4710794985294342,
"learning_rate": 0.00023121448768008613,
"loss": 2.9782,
"step": 91400
},
{
"epoch": 24.60992678725237,
"grad_norm": 0.4680551588535309,
"learning_rate": 0.00023101252187962838,
"loss": 2.9698,
"step": 91450
},
{
"epoch": 24.623385012919897,
"grad_norm": 0.47461768984794617,
"learning_rate": 0.00023081055607917058,
"loss": 2.9613,
"step": 91500
},
{
"epoch": 24.636843238587424,
"grad_norm": 0.42775940895080566,
"learning_rate": 0.0002306085902787128,
"loss": 2.9716,
"step": 91550
},
{
"epoch": 24.65030146425495,
"grad_norm": 0.491133451461792,
"learning_rate": 0.000230406624478255,
"loss": 2.9687,
"step": 91600
},
{
"epoch": 24.66375968992248,
"grad_norm": 0.43853914737701416,
"learning_rate": 0.0002302046586777972,
"loss": 2.9657,
"step": 91650
},
{
"epoch": 24.67721791559001,
"grad_norm": 0.46136704087257385,
"learning_rate": 0.00023000269287733942,
"loss": 2.9701,
"step": 91700
},
{
"epoch": 24.690676141257537,
"grad_norm": 0.5232241749763489,
"learning_rate": 0.00022980072707688161,
"loss": 2.9697,
"step": 91750
},
{
"epoch": 24.704134366925064,
"grad_norm": 0.45869165658950806,
"learning_rate": 0.00022959876127642384,
"loss": 2.9797,
"step": 91800
},
{
"epoch": 24.71759259259259,
"grad_norm": 0.46300241351127625,
"learning_rate": 0.00022939679547596603,
"loss": 2.9832,
"step": 91850
},
{
"epoch": 24.73105081826012,
"grad_norm": 0.456046462059021,
"learning_rate": 0.00022919482967550828,
"loss": 2.9793,
"step": 91900
},
{
"epoch": 24.74450904392765,
"grad_norm": 0.4491710066795349,
"learning_rate": 0.00022899286387505048,
"loss": 2.9819,
"step": 91950
},
{
"epoch": 24.757967269595177,
"grad_norm": 0.44479814171791077,
"learning_rate": 0.00022879089807459268,
"loss": 2.975,
"step": 92000
},
{
"epoch": 24.757967269595177,
"eval_accuracy": 0.39675495744737255,
"eval_loss": 3.2818639278411865,
"eval_runtime": 53.243,
"eval_samples_per_second": 338.298,
"eval_steps_per_second": 21.148,
"step": 92000
},
{
"epoch": 24.771425495262704,
"grad_norm": 0.4630224406719208,
"learning_rate": 0.0002285889322741349,
"loss": 2.9782,
"step": 92050
},
{
"epoch": 24.78488372093023,
"grad_norm": 0.4504377841949463,
"learning_rate": 0.0002283869664736771,
"loss": 2.9821,
"step": 92100
},
{
"epoch": 24.79834194659776,
"grad_norm": 0.46722739934921265,
"learning_rate": 0.00022818500067321932,
"loss": 2.9873,
"step": 92150
},
{
"epoch": 24.81180017226529,
"grad_norm": 0.4305804967880249,
"learning_rate": 0.00022798303487276151,
"loss": 2.9946,
"step": 92200
},
{
"epoch": 24.825258397932817,
"grad_norm": 0.46924400329589844,
"learning_rate": 0.0002277810690723037,
"loss": 2.9798,
"step": 92250
},
{
"epoch": 24.838716623600344,
"grad_norm": 0.4677594304084778,
"learning_rate": 0.00022757910327184596,
"loss": 2.9792,
"step": 92300
},
{
"epoch": 24.852174849267872,
"grad_norm": 0.4811413586139679,
"learning_rate": 0.00022737713747138818,
"loss": 2.9849,
"step": 92350
},
{
"epoch": 24.8656330749354,
"grad_norm": 0.4708091914653778,
"learning_rate": 0.00022717517167093038,
"loss": 2.9734,
"step": 92400
},
{
"epoch": 24.87909130060293,
"grad_norm": 0.4559333622455597,
"learning_rate": 0.00022697320587047258,
"loss": 2.9786,
"step": 92450
},
{
"epoch": 24.892549526270457,
"grad_norm": 0.4559614360332489,
"learning_rate": 0.0002267712400700148,
"loss": 2.9828,
"step": 92500
},
{
"epoch": 24.906007751937985,
"grad_norm": 0.4416915476322174,
"learning_rate": 0.000226569274269557,
"loss": 2.9819,
"step": 92550
},
{
"epoch": 24.919465977605512,
"grad_norm": 0.4463852047920227,
"learning_rate": 0.00022636730846909922,
"loss": 2.9988,
"step": 92600
},
{
"epoch": 24.93292420327304,
"grad_norm": 0.49153658747673035,
"learning_rate": 0.00022616534266864141,
"loss": 2.9787,
"step": 92650
},
{
"epoch": 24.94638242894057,
"grad_norm": 0.45088720321655273,
"learning_rate": 0.0002259633768681836,
"loss": 2.988,
"step": 92700
},
{
"epoch": 24.959840654608097,
"grad_norm": 0.4419504404067993,
"learning_rate": 0.00022576141106772586,
"loss": 2.978,
"step": 92750
},
{
"epoch": 24.973298880275625,
"grad_norm": 0.4792090058326721,
"learning_rate": 0.00022555944526726806,
"loss": 2.9729,
"step": 92800
},
{
"epoch": 24.986757105943152,
"grad_norm": 0.4608009159564972,
"learning_rate": 0.00022535747946681028,
"loss": 2.9824,
"step": 92850
},
{
"epoch": 25.0,
"grad_norm": 0.946661114692688,
"learning_rate": 0.00022515551366635248,
"loss": 2.9921,
"step": 92900
},
{
"epoch": 25.013458225667527,
"grad_norm": 0.46860387921333313,
"learning_rate": 0.0002249535478658947,
"loss": 2.9019,
"step": 92950
},
{
"epoch": 25.026916451335055,
"grad_norm": 0.4779540002346039,
"learning_rate": 0.0002247515820654369,
"loss": 2.8962,
"step": 93000
},
{
"epoch": 25.026916451335055,
"eval_accuracy": 0.3965091982600858,
"eval_loss": 3.290764331817627,
"eval_runtime": 53.1111,
"eval_samples_per_second": 339.138,
"eval_steps_per_second": 21.201,
"step": 93000
},
{
"epoch": 25.040374677002585,
"grad_norm": 0.4737008810043335,
"learning_rate": 0.0002245496162649791,
"loss": 2.9012,
"step": 93050
},
{
"epoch": 25.053832902670113,
"grad_norm": 0.45304667949676514,
"learning_rate": 0.00022434765046452131,
"loss": 2.9132,
"step": 93100
},
{
"epoch": 25.06729112833764,
"grad_norm": 0.49432122707366943,
"learning_rate": 0.0002241456846640635,
"loss": 2.9148,
"step": 93150
},
{
"epoch": 25.080749354005167,
"grad_norm": 0.4715327024459839,
"learning_rate": 0.00022394371886360576,
"loss": 2.9212,
"step": 93200
},
{
"epoch": 25.094207579672695,
"grad_norm": 0.4631378650665283,
"learning_rate": 0.00022374175306314796,
"loss": 2.9205,
"step": 93250
},
{
"epoch": 25.107665805340226,
"grad_norm": 0.44028857350349426,
"learning_rate": 0.00022353978726269018,
"loss": 2.9289,
"step": 93300
},
{
"epoch": 25.121124031007753,
"grad_norm": 0.481253981590271,
"learning_rate": 0.00022333782146223238,
"loss": 2.9248,
"step": 93350
},
{
"epoch": 25.13458225667528,
"grad_norm": 0.47388628125190735,
"learning_rate": 0.0002231358556617746,
"loss": 2.9136,
"step": 93400
},
{
"epoch": 25.148040482342807,
"grad_norm": 0.471007376909256,
"learning_rate": 0.0002229338898613168,
"loss": 2.9299,
"step": 93450
},
{
"epoch": 25.161498708010335,
"grad_norm": 0.47656720876693726,
"learning_rate": 0.000222731924060859,
"loss": 2.9299,
"step": 93500
},
{
"epoch": 25.174956933677866,
"grad_norm": 0.48185816407203674,
"learning_rate": 0.00022252995826040122,
"loss": 2.9189,
"step": 93550
},
{
"epoch": 25.188415159345393,
"grad_norm": 0.4834353029727936,
"learning_rate": 0.0002223279924599434,
"loss": 2.9224,
"step": 93600
},
{
"epoch": 25.20187338501292,
"grad_norm": 0.45164158940315247,
"learning_rate": 0.00022212602665948566,
"loss": 2.9373,
"step": 93650
},
{
"epoch": 25.215331610680447,
"grad_norm": 0.4696139693260193,
"learning_rate": 0.00022192406085902786,
"loss": 2.9385,
"step": 93700
},
{
"epoch": 25.228789836347975,
"grad_norm": 0.45761996507644653,
"learning_rate": 0.00022172209505857008,
"loss": 2.9435,
"step": 93750
},
{
"epoch": 25.242248062015506,
"grad_norm": 0.4780551493167877,
"learning_rate": 0.00022152012925811228,
"loss": 2.9396,
"step": 93800
},
{
"epoch": 25.255706287683033,
"grad_norm": 0.4723582863807678,
"learning_rate": 0.00022131816345765447,
"loss": 2.9307,
"step": 93850
},
{
"epoch": 25.26916451335056,
"grad_norm": 0.4635056257247925,
"learning_rate": 0.0002211161976571967,
"loss": 2.9335,
"step": 93900
},
{
"epoch": 25.282622739018088,
"grad_norm": 0.4594542980194092,
"learning_rate": 0.0002209142318567389,
"loss": 2.9302,
"step": 93950
},
{
"epoch": 25.296080964685615,
"grad_norm": 0.45935070514678955,
"learning_rate": 0.00022071226605628112,
"loss": 2.9417,
"step": 94000
},
{
"epoch": 25.296080964685615,
"eval_accuracy": 0.39655493858627844,
"eval_loss": 3.287743330001831,
"eval_runtime": 53.1973,
"eval_samples_per_second": 338.589,
"eval_steps_per_second": 21.166,
"step": 94000
},
{
"epoch": 25.309539190353142,
"grad_norm": 0.4579530358314514,
"learning_rate": 0.0002205103002558233,
"loss": 2.9377,
"step": 94050
},
{
"epoch": 25.322997416020673,
"grad_norm": 0.5007568597793579,
"learning_rate": 0.00022030833445536556,
"loss": 2.9477,
"step": 94100
},
{
"epoch": 25.3364556416882,
"grad_norm": 0.47446051239967346,
"learning_rate": 0.00022010636865490776,
"loss": 2.9362,
"step": 94150
},
{
"epoch": 25.349913867355728,
"grad_norm": 0.444487065076828,
"learning_rate": 0.00021990440285444998,
"loss": 2.9374,
"step": 94200
},
{
"epoch": 25.363372093023255,
"grad_norm": 0.4731683135032654,
"learning_rate": 0.00021970243705399218,
"loss": 2.9463,
"step": 94250
},
{
"epoch": 25.376830318690782,
"grad_norm": 0.45736849308013916,
"learning_rate": 0.00021950047125353437,
"loss": 2.9414,
"step": 94300
},
{
"epoch": 25.390288544358313,
"grad_norm": 0.45242130756378174,
"learning_rate": 0.0002192985054530766,
"loss": 2.9547,
"step": 94350
},
{
"epoch": 25.40374677002584,
"grad_norm": 0.45447608828544617,
"learning_rate": 0.0002190965396526188,
"loss": 2.9417,
"step": 94400
},
{
"epoch": 25.417204995693368,
"grad_norm": 0.46485042572021484,
"learning_rate": 0.00021889457385216102,
"loss": 2.9464,
"step": 94450
},
{
"epoch": 25.430663221360895,
"grad_norm": 0.45448994636535645,
"learning_rate": 0.0002186926080517032,
"loss": 2.9645,
"step": 94500
},
{
"epoch": 25.444121447028422,
"grad_norm": 0.5003591179847717,
"learning_rate": 0.00021849064225124546,
"loss": 2.9631,
"step": 94550
},
{
"epoch": 25.457579672695953,
"grad_norm": 0.466730535030365,
"learning_rate": 0.00021828867645078766,
"loss": 2.9466,
"step": 94600
},
{
"epoch": 25.47103789836348,
"grad_norm": 0.4516577124595642,
"learning_rate": 0.00021808671065032985,
"loss": 2.9398,
"step": 94650
},
{
"epoch": 25.484496124031008,
"grad_norm": 0.47522616386413574,
"learning_rate": 0.00021788474484987208,
"loss": 2.9451,
"step": 94700
},
{
"epoch": 25.497954349698535,
"grad_norm": 0.5053650140762329,
"learning_rate": 0.00021768277904941427,
"loss": 2.9583,
"step": 94750
},
{
"epoch": 25.511412575366062,
"grad_norm": 0.5291735529899597,
"learning_rate": 0.0002174808132489565,
"loss": 2.9607,
"step": 94800
},
{
"epoch": 25.524870801033593,
"grad_norm": 0.4376024007797241,
"learning_rate": 0.0002172788474484987,
"loss": 2.9555,
"step": 94850
},
{
"epoch": 25.53832902670112,
"grad_norm": 0.4744149148464203,
"learning_rate": 0.0002170768816480409,
"loss": 2.9574,
"step": 94900
},
{
"epoch": 25.551787252368648,
"grad_norm": 0.46446311473846436,
"learning_rate": 0.0002168749158475831,
"loss": 2.9484,
"step": 94950
},
{
"epoch": 25.565245478036175,
"grad_norm": 0.5329943895339966,
"learning_rate": 0.00021667295004712534,
"loss": 2.9678,
"step": 95000
},
{
"epoch": 25.565245478036175,
"eval_accuracy": 0.3970521065973413,
"eval_loss": 3.284172773361206,
"eval_runtime": 53.2481,
"eval_samples_per_second": 338.265,
"eval_steps_per_second": 21.146,
"step": 95000
},
{
"epoch": 25.578703703703702,
"grad_norm": 0.45800742506980896,
"learning_rate": 0.00021647098424666756,
"loss": 2.9489,
"step": 95050
},
{
"epoch": 25.592161929371233,
"grad_norm": 0.4492628872394562,
"learning_rate": 0.00021626901844620976,
"loss": 2.9634,
"step": 95100
},
{
"epoch": 25.60562015503876,
"grad_norm": 0.4850592613220215,
"learning_rate": 0.00021606705264575198,
"loss": 2.9495,
"step": 95150
},
{
"epoch": 25.619078380706288,
"grad_norm": 0.4660792946815491,
"learning_rate": 0.00021586508684529417,
"loss": 2.9646,
"step": 95200
},
{
"epoch": 25.632536606373815,
"grad_norm": 0.5027803778648376,
"learning_rate": 0.0002156631210448364,
"loss": 2.9722,
"step": 95250
},
{
"epoch": 25.645994832041342,
"grad_norm": 0.4320453703403473,
"learning_rate": 0.0002154611552443786,
"loss": 2.9601,
"step": 95300
},
{
"epoch": 25.659453057708873,
"grad_norm": 0.507905900478363,
"learning_rate": 0.0002152591894439208,
"loss": 2.9649,
"step": 95350
},
{
"epoch": 25.6729112833764,
"grad_norm": 0.4640471637248993,
"learning_rate": 0.000215057223643463,
"loss": 2.96,
"step": 95400
},
{
"epoch": 25.686369509043928,
"grad_norm": 0.4867199659347534,
"learning_rate": 0.00021485525784300524,
"loss": 2.9652,
"step": 95450
},
{
"epoch": 25.699827734711455,
"grad_norm": 0.5132190585136414,
"learning_rate": 0.00021465329204254746,
"loss": 2.9669,
"step": 95500
},
{
"epoch": 25.713285960378983,
"grad_norm": 0.4502236247062683,
"learning_rate": 0.00021445132624208966,
"loss": 2.9701,
"step": 95550
},
{
"epoch": 25.72674418604651,
"grad_norm": 0.5049893856048584,
"learning_rate": 0.00021424936044163188,
"loss": 2.9641,
"step": 95600
},
{
"epoch": 25.74020241171404,
"grad_norm": 0.4590776860713959,
"learning_rate": 0.00021404739464117407,
"loss": 2.9627,
"step": 95650
},
{
"epoch": 25.753660637381568,
"grad_norm": 0.46981340646743774,
"learning_rate": 0.00021384542884071627,
"loss": 2.9816,
"step": 95700
},
{
"epoch": 25.767118863049095,
"grad_norm": 0.4755721688270569,
"learning_rate": 0.0002136434630402585,
"loss": 2.981,
"step": 95750
},
{
"epoch": 25.780577088716623,
"grad_norm": 0.4739704132080078,
"learning_rate": 0.0002134414972398007,
"loss": 2.9738,
"step": 95800
},
{
"epoch": 25.794035314384153,
"grad_norm": 0.46069255471229553,
"learning_rate": 0.0002132395314393429,
"loss": 2.9837,
"step": 95850
},
{
"epoch": 25.80749354005168,
"grad_norm": 0.4511219561100006,
"learning_rate": 0.00021303756563888514,
"loss": 2.9805,
"step": 95900
},
{
"epoch": 25.820951765719208,
"grad_norm": 0.4511417746543884,
"learning_rate": 0.00021283559983842736,
"loss": 2.9704,
"step": 95950
},
{
"epoch": 25.834409991386735,
"grad_norm": 0.45782825350761414,
"learning_rate": 0.00021263363403796956,
"loss": 2.9681,
"step": 96000
},
{
"epoch": 25.834409991386735,
"eval_accuracy": 0.3971620572151725,
"eval_loss": 3.2770164012908936,
"eval_runtime": 53.2448,
"eval_samples_per_second": 338.287,
"eval_steps_per_second": 21.148,
"step": 96000
},
{
"epoch": 25.847868217054263,
"grad_norm": 0.46823206543922424,
"learning_rate": 0.00021243166823751175,
"loss": 2.9675,
"step": 96050
},
{
"epoch": 25.86132644272179,
"grad_norm": 0.5006187558174133,
"learning_rate": 0.00021222970243705398,
"loss": 2.9624,
"step": 96100
},
{
"epoch": 25.87478466838932,
"grad_norm": 0.47216886281967163,
"learning_rate": 0.00021202773663659617,
"loss": 2.9821,
"step": 96150
},
{
"epoch": 25.888242894056848,
"grad_norm": 0.46271994709968567,
"learning_rate": 0.0002118257708361384,
"loss": 2.9807,
"step": 96200
},
{
"epoch": 25.901701119724375,
"grad_norm": 0.49749988317489624,
"learning_rate": 0.0002116238050356806,
"loss": 2.9749,
"step": 96250
},
{
"epoch": 25.915159345391903,
"grad_norm": 0.4832835793495178,
"learning_rate": 0.00021142183923522281,
"loss": 2.9753,
"step": 96300
},
{
"epoch": 25.92861757105943,
"grad_norm": 0.49002805352211,
"learning_rate": 0.00021121987343476504,
"loss": 2.9831,
"step": 96350
},
{
"epoch": 25.94207579672696,
"grad_norm": 0.466495156288147,
"learning_rate": 0.00021101790763430726,
"loss": 2.9786,
"step": 96400
},
{
"epoch": 25.955534022394488,
"grad_norm": 0.4698829650878906,
"learning_rate": 0.00021081594183384946,
"loss": 2.9683,
"step": 96450
},
{
"epoch": 25.968992248062015,
"grad_norm": 0.46221068501472473,
"learning_rate": 0.00021061397603339165,
"loss": 2.9681,
"step": 96500
},
{
"epoch": 25.982450473729543,
"grad_norm": 0.4581020772457123,
"learning_rate": 0.00021041201023293388,
"loss": 2.981,
"step": 96550
},
{
"epoch": 25.99590869939707,
"grad_norm": 0.4899604022502899,
"learning_rate": 0.00021021004443247607,
"loss": 2.9846,
"step": 96600
},
{
"epoch": 26.009151593453918,
"grad_norm": 0.4719387888908386,
"learning_rate": 0.0002100080786320183,
"loss": 2.9147,
"step": 96650
},
{
"epoch": 26.022609819121445,
"grad_norm": 0.4555143713951111,
"learning_rate": 0.0002098061128315605,
"loss": 2.8874,
"step": 96700
},
{
"epoch": 26.036068044788976,
"grad_norm": 0.46729525923728943,
"learning_rate": 0.00020960414703110274,
"loss": 2.8964,
"step": 96750
},
{
"epoch": 26.049526270456504,
"grad_norm": 0.45050865411758423,
"learning_rate": 0.00020940218123064494,
"loss": 2.9157,
"step": 96800
},
{
"epoch": 26.06298449612403,
"grad_norm": 0.4598424434661865,
"learning_rate": 0.00020920021543018713,
"loss": 2.908,
"step": 96850
},
{
"epoch": 26.076442721791558,
"grad_norm": 0.4923440217971802,
"learning_rate": 0.00020899824962972936,
"loss": 2.9066,
"step": 96900
},
{
"epoch": 26.089900947459086,
"grad_norm": 0.46300119161605835,
"learning_rate": 0.00020879628382927155,
"loss": 2.9051,
"step": 96950
},
{
"epoch": 26.103359173126616,
"grad_norm": 0.4792136549949646,
"learning_rate": 0.00020859431802881378,
"loss": 2.9053,
"step": 97000
},
{
"epoch": 26.103359173126616,
"eval_accuracy": 0.3962396454114304,
"eval_loss": 3.2915122509002686,
"eval_runtime": 53.217,
"eval_samples_per_second": 338.463,
"eval_steps_per_second": 21.159,
"step": 97000
},
{
"epoch": 26.116817398794144,
"grad_norm": 0.45519232749938965,
"learning_rate": 0.00020839235222835597,
"loss": 2.916,
"step": 97050
},
{
"epoch": 26.13027562446167,
"grad_norm": 0.47248655557632446,
"learning_rate": 0.00020819038642789817,
"loss": 2.9061,
"step": 97100
},
{
"epoch": 26.1437338501292,
"grad_norm": 0.4677557051181793,
"learning_rate": 0.0002079884206274404,
"loss": 2.925,
"step": 97150
},
{
"epoch": 26.157192075796726,
"grad_norm": 0.4564340114593506,
"learning_rate": 0.00020778645482698264,
"loss": 2.9118,
"step": 97200
},
{
"epoch": 26.170650301464256,
"grad_norm": 0.47288423776626587,
"learning_rate": 0.00020758448902652484,
"loss": 2.9354,
"step": 97250
},
{
"epoch": 26.184108527131784,
"grad_norm": 0.5112426280975342,
"learning_rate": 0.00020738252322606703,
"loss": 2.913,
"step": 97300
},
{
"epoch": 26.19756675279931,
"grad_norm": 0.48271527886390686,
"learning_rate": 0.00020718055742560926,
"loss": 2.9189,
"step": 97350
},
{
"epoch": 26.21102497846684,
"grad_norm": 0.4612698256969452,
"learning_rate": 0.00020697859162515145,
"loss": 2.9164,
"step": 97400
},
{
"epoch": 26.224483204134366,
"grad_norm": 0.48598477244377136,
"learning_rate": 0.00020677662582469368,
"loss": 2.9385,
"step": 97450
},
{
"epoch": 26.237941429801896,
"grad_norm": 0.4755497872829437,
"learning_rate": 0.00020657466002423587,
"loss": 2.9334,
"step": 97500
},
{
"epoch": 26.251399655469424,
"grad_norm": 0.4690141975879669,
"learning_rate": 0.00020637269422377807,
"loss": 2.9302,
"step": 97550
},
{
"epoch": 26.26485788113695,
"grad_norm": 0.45227643847465515,
"learning_rate": 0.0002061707284233203,
"loss": 2.9261,
"step": 97600
},
{
"epoch": 26.27831610680448,
"grad_norm": 0.4535977244377136,
"learning_rate": 0.00020596876262286251,
"loss": 2.9256,
"step": 97650
},
{
"epoch": 26.291774332472006,
"grad_norm": 0.46246781945228577,
"learning_rate": 0.00020576679682240474,
"loss": 2.9267,
"step": 97700
},
{
"epoch": 26.305232558139537,
"grad_norm": 0.47968053817749023,
"learning_rate": 0.00020556483102194693,
"loss": 2.9305,
"step": 97750
},
{
"epoch": 26.318690783807064,
"grad_norm": 0.45259711146354675,
"learning_rate": 0.00020536286522148916,
"loss": 2.9192,
"step": 97800
},
{
"epoch": 26.33214900947459,
"grad_norm": 0.45785126090049744,
"learning_rate": 0.00020516089942103135,
"loss": 2.9314,
"step": 97850
},
{
"epoch": 26.34560723514212,
"grad_norm": 0.4738333821296692,
"learning_rate": 0.00020495893362057355,
"loss": 2.9313,
"step": 97900
},
{
"epoch": 26.359065460809646,
"grad_norm": 0.47246575355529785,
"learning_rate": 0.00020475696782011577,
"loss": 2.937,
"step": 97950
},
{
"epoch": 26.372523686477177,
"grad_norm": 0.5067406892776489,
"learning_rate": 0.00020455500201965797,
"loss": 2.9413,
"step": 98000
},
{
"epoch": 26.372523686477177,
"eval_accuracy": 0.3968445911032132,
"eval_loss": 3.289306163787842,
"eval_runtime": 53.1207,
"eval_samples_per_second": 339.077,
"eval_steps_per_second": 21.197,
"step": 98000
},
{
"epoch": 26.385981912144704,
"grad_norm": 0.4930259883403778,
"learning_rate": 0.0002043530362192002,
"loss": 2.9361,
"step": 98050
},
{
"epoch": 26.39944013781223,
"grad_norm": 0.48024407029151917,
"learning_rate": 0.00020415107041874242,
"loss": 2.9505,
"step": 98100
},
{
"epoch": 26.41289836347976,
"grad_norm": 0.47633829712867737,
"learning_rate": 0.00020394910461828464,
"loss": 2.9337,
"step": 98150
},
{
"epoch": 26.426356589147286,
"grad_norm": 0.48809248208999634,
"learning_rate": 0.00020374713881782683,
"loss": 2.9409,
"step": 98200
},
{
"epoch": 26.439814814814813,
"grad_norm": 0.4671519100666046,
"learning_rate": 0.00020354517301736906,
"loss": 2.946,
"step": 98250
},
{
"epoch": 26.453273040482344,
"grad_norm": 0.5086513757705688,
"learning_rate": 0.00020334320721691125,
"loss": 2.9288,
"step": 98300
},
{
"epoch": 26.46673126614987,
"grad_norm": 0.4985395073890686,
"learning_rate": 0.00020314124141645345,
"loss": 2.9412,
"step": 98350
},
{
"epoch": 26.4801894918174,
"grad_norm": 0.5244379639625549,
"learning_rate": 0.00020293927561599567,
"loss": 2.9527,
"step": 98400
},
{
"epoch": 26.493647717484926,
"grad_norm": 0.4727938175201416,
"learning_rate": 0.00020273730981553787,
"loss": 2.9467,
"step": 98450
},
{
"epoch": 26.507105943152453,
"grad_norm": 0.4935176372528076,
"learning_rate": 0.0002025353440150801,
"loss": 2.9375,
"step": 98500
},
{
"epoch": 26.520564168819984,
"grad_norm": 0.4935429096221924,
"learning_rate": 0.00020233337821462232,
"loss": 2.9595,
"step": 98550
},
{
"epoch": 26.53402239448751,
"grad_norm": 0.49508944153785706,
"learning_rate": 0.00020213141241416454,
"loss": 2.9458,
"step": 98600
},
{
"epoch": 26.54748062015504,
"grad_norm": 0.5003538131713867,
"learning_rate": 0.00020192944661370674,
"loss": 2.9514,
"step": 98650
},
{
"epoch": 26.560938845822566,
"grad_norm": 0.48732179403305054,
"learning_rate": 0.00020172748081324893,
"loss": 2.9505,
"step": 98700
},
{
"epoch": 26.574397071490093,
"grad_norm": 0.4849097430706024,
"learning_rate": 0.00020152551501279115,
"loss": 2.9532,
"step": 98750
},
{
"epoch": 26.587855297157624,
"grad_norm": 0.4961356222629547,
"learning_rate": 0.00020132354921233335,
"loss": 2.9481,
"step": 98800
},
{
"epoch": 26.60131352282515,
"grad_norm": 0.47980841994285583,
"learning_rate": 0.00020112158341187557,
"loss": 2.9566,
"step": 98850
},
{
"epoch": 26.61477174849268,
"grad_norm": 0.4506068825721741,
"learning_rate": 0.00020091961761141777,
"loss": 2.9509,
"step": 98900
},
{
"epoch": 26.628229974160206,
"grad_norm": 0.47053366899490356,
"learning_rate": 0.00020071765181095997,
"loss": 2.9551,
"step": 98950
},
{
"epoch": 26.641688199827733,
"grad_norm": 0.45985883474349976,
"learning_rate": 0.00020051568601050222,
"loss": 2.9376,
"step": 99000
},
{
"epoch": 26.641688199827733,
"eval_accuracy": 0.39680080642042076,
"eval_loss": 3.2819876670837402,
"eval_runtime": 53.2269,
"eval_samples_per_second": 338.4,
"eval_steps_per_second": 21.155,
"step": 99000
},
{
"epoch": 26.655146425495264,
"grad_norm": 0.45460259914398193,
"learning_rate": 0.00020031372021004444,
"loss": 2.9569,
"step": 99050
},
{
"epoch": 26.66860465116279,
"grad_norm": 0.4597529470920563,
"learning_rate": 0.00020011175440958664,
"loss": 2.9559,
"step": 99100
},
{
"epoch": 26.68206287683032,
"grad_norm": 0.4680781364440918,
"learning_rate": 0.00019990978860912883,
"loss": 2.9514,
"step": 99150
},
{
"epoch": 26.695521102497846,
"grad_norm": 0.4671817123889923,
"learning_rate": 0.00019970782280867105,
"loss": 2.9426,
"step": 99200
},
{
"epoch": 26.708979328165373,
"grad_norm": 0.4486178159713745,
"learning_rate": 0.00019950585700821325,
"loss": 2.9519,
"step": 99250
},
{
"epoch": 26.722437553832904,
"grad_norm": 0.47057074308395386,
"learning_rate": 0.00019930389120775547,
"loss": 2.9607,
"step": 99300
},
{
"epoch": 26.73589577950043,
"grad_norm": 0.45896413922309875,
"learning_rate": 0.00019910192540729767,
"loss": 2.9562,
"step": 99350
},
{
"epoch": 26.74935400516796,
"grad_norm": 0.48428693413734436,
"learning_rate": 0.00019889995960683987,
"loss": 2.955,
"step": 99400
},
{
"epoch": 26.762812230835486,
"grad_norm": 0.4974263608455658,
"learning_rate": 0.00019869799380638212,
"loss": 2.9546,
"step": 99450
},
{
"epoch": 26.776270456503013,
"grad_norm": 0.49993783235549927,
"learning_rate": 0.0001984960280059243,
"loss": 2.9653,
"step": 99500
},
{
"epoch": 26.789728682170544,
"grad_norm": 0.5132726430892944,
"learning_rate": 0.00019829406220546654,
"loss": 2.9621,
"step": 99550
},
{
"epoch": 26.80318690783807,
"grad_norm": 0.4783254563808441,
"learning_rate": 0.00019809209640500873,
"loss": 2.9638,
"step": 99600
},
{
"epoch": 26.8166451335056,
"grad_norm": 0.5352595448493958,
"learning_rate": 0.00019789013060455096,
"loss": 2.9615,
"step": 99650
},
{
"epoch": 26.830103359173126,
"grad_norm": 0.49175021052360535,
"learning_rate": 0.00019768816480409315,
"loss": 2.968,
"step": 99700
},
{
"epoch": 26.843561584840653,
"grad_norm": 0.5543900728225708,
"learning_rate": 0.00019748619900363535,
"loss": 2.9575,
"step": 99750
},
{
"epoch": 26.857019810508184,
"grad_norm": 0.4724408686161041,
"learning_rate": 0.00019728423320317757,
"loss": 2.956,
"step": 99800
},
{
"epoch": 26.87047803617571,
"grad_norm": 0.4760195016860962,
"learning_rate": 0.00019708226740271977,
"loss": 2.9631,
"step": 99850
},
{
"epoch": 26.88393626184324,
"grad_norm": 0.46866995096206665,
"learning_rate": 0.00019688030160226202,
"loss": 2.9725,
"step": 99900
},
{
"epoch": 26.897394487510766,
"grad_norm": 0.4831426739692688,
"learning_rate": 0.0001966783358018042,
"loss": 2.9608,
"step": 99950
},
{
"epoch": 26.910852713178294,
"grad_norm": 0.4839981198310852,
"learning_rate": 0.00019647637000134644,
"loss": 2.971,
"step": 100000
},
{
"epoch": 26.910852713178294,
"eval_accuracy": 0.3974456255081957,
"eval_loss": 3.278042793273926,
"eval_runtime": 53.2475,
"eval_samples_per_second": 338.27,
"eval_steps_per_second": 21.147,
"step": 100000
},
{
"epoch": 26.92431093884582,
"grad_norm": 0.4651424288749695,
"learning_rate": 0.00019627440420088863,
"loss": 2.9675,
"step": 100050
},
{
"epoch": 26.93776916451335,
"grad_norm": 0.4988207519054413,
"learning_rate": 0.00019607243840043086,
"loss": 2.9717,
"step": 100100
},
{
"epoch": 26.95122739018088,
"grad_norm": 0.4863513112068176,
"learning_rate": 0.00019587047259997305,
"loss": 2.9627,
"step": 100150
},
{
"epoch": 26.964685615848406,
"grad_norm": 0.494176983833313,
"learning_rate": 0.00019566850679951525,
"loss": 2.9718,
"step": 100200
},
{
"epoch": 26.978143841515934,
"grad_norm": 0.529505729675293,
"learning_rate": 0.00019546654099905747,
"loss": 2.9744,
"step": 100250
},
{
"epoch": 26.99160206718346,
"grad_norm": 0.5348026752471924,
"learning_rate": 0.00019526457519859967,
"loss": 2.9587,
"step": 100300
},
{
"epoch": 27.00484496124031,
"grad_norm": 0.4647117555141449,
"learning_rate": 0.00019506260939814192,
"loss": 2.9395,
"step": 100350
},
{
"epoch": 27.01830318690784,
"grad_norm": 0.4926200807094574,
"learning_rate": 0.0001948606435976841,
"loss": 2.897,
"step": 100400
},
{
"epoch": 27.031761412575367,
"grad_norm": 0.5165723562240601,
"learning_rate": 0.00019465867779722634,
"loss": 2.8961,
"step": 100450
},
{
"epoch": 27.045219638242894,
"grad_norm": 0.4593856632709503,
"learning_rate": 0.00019445671199676853,
"loss": 2.8818,
"step": 100500
},
{
"epoch": 27.05867786391042,
"grad_norm": 0.5317386388778687,
"learning_rate": 0.00019425474619631073,
"loss": 2.8978,
"step": 100550
},
{
"epoch": 27.07213608957795,
"grad_norm": 0.5097870230674744,
"learning_rate": 0.00019405278039585295,
"loss": 2.8983,
"step": 100600
},
{
"epoch": 27.085594315245476,
"grad_norm": 0.5453401803970337,
"learning_rate": 0.00019385081459539515,
"loss": 2.8999,
"step": 100650
},
{
"epoch": 27.099052540913007,
"grad_norm": 0.4838346540927887,
"learning_rate": 0.00019364884879493737,
"loss": 2.8915,
"step": 100700
},
{
"epoch": 27.112510766580534,
"grad_norm": 0.4938969612121582,
"learning_rate": 0.00019344688299447957,
"loss": 2.9024,
"step": 100750
},
{
"epoch": 27.125968992248062,
"grad_norm": 0.4785228669643402,
"learning_rate": 0.00019324491719402182,
"loss": 2.9025,
"step": 100800
},
{
"epoch": 27.13942721791559,
"grad_norm": 0.48732790350914,
"learning_rate": 0.00019304295139356401,
"loss": 2.8985,
"step": 100850
},
{
"epoch": 27.152885443583116,
"grad_norm": 0.48145633935928345,
"learning_rate": 0.0001928409855931062,
"loss": 2.9025,
"step": 100900
},
{
"epoch": 27.166343669250647,
"grad_norm": 0.48346710205078125,
"learning_rate": 0.00019263901979264843,
"loss": 2.9107,
"step": 100950
},
{
"epoch": 27.179801894918175,
"grad_norm": 0.46865689754486084,
"learning_rate": 0.00019243705399219063,
"loss": 2.9114,
"step": 101000
},
{
"epoch": 27.179801894918175,
"eval_accuracy": 0.3968449170437799,
"eval_loss": 3.2898991107940674,
"eval_runtime": 53.2817,
"eval_samples_per_second": 338.052,
"eval_steps_per_second": 21.133,
"step": 101000
},
{
"epoch": 27.193260120585702,
"grad_norm": 0.5351729393005371,
"learning_rate": 0.00019223508819173285,
"loss": 2.9076,
"step": 101050
},
{
"epoch": 27.20671834625323,
"grad_norm": 0.4565810263156891,
"learning_rate": 0.00019203312239127505,
"loss": 2.919,
"step": 101100
},
{
"epoch": 27.220176571920756,
"grad_norm": 0.4939174950122833,
"learning_rate": 0.00019183115659081727,
"loss": 2.9204,
"step": 101150
},
{
"epoch": 27.233634797588287,
"grad_norm": 0.5174657702445984,
"learning_rate": 0.00019162919079035947,
"loss": 2.9102,
"step": 101200
},
{
"epoch": 27.247093023255815,
"grad_norm": 0.4990021586418152,
"learning_rate": 0.00019142722498990172,
"loss": 2.9231,
"step": 101250
},
{
"epoch": 27.260551248923342,
"grad_norm": 0.5200758576393127,
"learning_rate": 0.00019122525918944391,
"loss": 2.9277,
"step": 101300
},
{
"epoch": 27.27400947459087,
"grad_norm": 0.5058716535568237,
"learning_rate": 0.0001910232933889861,
"loss": 2.9167,
"step": 101350
},
{
"epoch": 27.287467700258397,
"grad_norm": 0.47531625628471375,
"learning_rate": 0.00019082132758852833,
"loss": 2.9179,
"step": 101400
},
{
"epoch": 27.300925925925927,
"grad_norm": 0.4710346460342407,
"learning_rate": 0.00019061936178807053,
"loss": 2.9201,
"step": 101450
},
{
"epoch": 27.314384151593455,
"grad_norm": 0.4861701726913452,
"learning_rate": 0.00019041739598761275,
"loss": 2.9216,
"step": 101500
},
{
"epoch": 27.327842377260982,
"grad_norm": 0.4996824562549591,
"learning_rate": 0.00019021543018715495,
"loss": 2.9229,
"step": 101550
},
{
"epoch": 27.34130060292851,
"grad_norm": 0.5007565021514893,
"learning_rate": 0.00019001346438669714,
"loss": 2.9237,
"step": 101600
},
{
"epoch": 27.354758828596037,
"grad_norm": 0.4790315628051758,
"learning_rate": 0.0001898114985862394,
"loss": 2.9219,
"step": 101650
},
{
"epoch": 27.368217054263567,
"grad_norm": 0.4641251862049103,
"learning_rate": 0.0001896095327857816,
"loss": 2.9413,
"step": 101700
},
{
"epoch": 27.381675279931095,
"grad_norm": 0.48673513531684875,
"learning_rate": 0.00018940756698532381,
"loss": 2.9245,
"step": 101750
},
{
"epoch": 27.395133505598622,
"grad_norm": 0.4806927740573883,
"learning_rate": 0.000189205601184866,
"loss": 2.9264,
"step": 101800
},
{
"epoch": 27.40859173126615,
"grad_norm": 0.48427829146385193,
"learning_rate": 0.00018900363538440823,
"loss": 2.9298,
"step": 101850
},
{
"epoch": 27.422049956933677,
"grad_norm": 0.5548421740531921,
"learning_rate": 0.00018880166958395043,
"loss": 2.9346,
"step": 101900
},
{
"epoch": 27.435508182601207,
"grad_norm": 0.4557640254497528,
"learning_rate": 0.00018859970378349263,
"loss": 2.925,
"step": 101950
},
{
"epoch": 27.448966408268735,
"grad_norm": 0.5101822018623352,
"learning_rate": 0.00018839773798303485,
"loss": 2.9329,
"step": 102000
},
{
"epoch": 27.448966408268735,
"eval_accuracy": 0.3970826363637549,
"eval_loss": 3.2858848571777344,
"eval_runtime": 53.2816,
"eval_samples_per_second": 338.053,
"eval_steps_per_second": 21.133,
"step": 102000
},
{
"epoch": 27.462424633936262,
"grad_norm": 0.4724166989326477,
"learning_rate": 0.00018819577218257705,
"loss": 2.9356,
"step": 102050
},
{
"epoch": 27.47588285960379,
"grad_norm": 0.4706898033618927,
"learning_rate": 0.0001879938063821193,
"loss": 2.9316,
"step": 102100
},
{
"epoch": 27.489341085271317,
"grad_norm": 0.50807124376297,
"learning_rate": 0.0001877918405816615,
"loss": 2.9286,
"step": 102150
},
{
"epoch": 27.502799310938848,
"grad_norm": 0.49092021584510803,
"learning_rate": 0.00018758987478120371,
"loss": 2.9353,
"step": 102200
},
{
"epoch": 27.516257536606375,
"grad_norm": 0.487447589635849,
"learning_rate": 0.0001873879089807459,
"loss": 2.9379,
"step": 102250
},
{
"epoch": 27.529715762273902,
"grad_norm": 0.47301945090293884,
"learning_rate": 0.00018718594318028813,
"loss": 2.9333,
"step": 102300
},
{
"epoch": 27.54317398794143,
"grad_norm": 0.4922998547554016,
"learning_rate": 0.00018698397737983033,
"loss": 2.9339,
"step": 102350
},
{
"epoch": 27.556632213608957,
"grad_norm": 0.46500828862190247,
"learning_rate": 0.00018678201157937253,
"loss": 2.9441,
"step": 102400
},
{
"epoch": 27.570090439276484,
"grad_norm": 0.4921210706233978,
"learning_rate": 0.00018658004577891475,
"loss": 2.9469,
"step": 102450
},
{
"epoch": 27.583548664944015,
"grad_norm": 0.4869675040245056,
"learning_rate": 0.00018637807997845695,
"loss": 2.9387,
"step": 102500
},
{
"epoch": 27.597006890611542,
"grad_norm": 0.46710294485092163,
"learning_rate": 0.0001861761141779992,
"loss": 2.9361,
"step": 102550
},
{
"epoch": 27.61046511627907,
"grad_norm": 0.5193657279014587,
"learning_rate": 0.0001859741483775414,
"loss": 2.93,
"step": 102600
},
{
"epoch": 27.623923341946597,
"grad_norm": 0.48783352971076965,
"learning_rate": 0.00018577218257708362,
"loss": 2.9433,
"step": 102650
},
{
"epoch": 27.637381567614124,
"grad_norm": 0.4768672585487366,
"learning_rate": 0.0001855702167766258,
"loss": 2.9442,
"step": 102700
},
{
"epoch": 27.650839793281655,
"grad_norm": 0.5465898513793945,
"learning_rate": 0.000185368250976168,
"loss": 2.9463,
"step": 102750
},
{
"epoch": 27.664298018949182,
"grad_norm": 0.48885899782180786,
"learning_rate": 0.00018516628517571023,
"loss": 2.9421,
"step": 102800
},
{
"epoch": 27.67775624461671,
"grad_norm": 0.48874956369400024,
"learning_rate": 0.00018496431937525243,
"loss": 2.94,
"step": 102850
},
{
"epoch": 27.691214470284237,
"grad_norm": 0.4969959259033203,
"learning_rate": 0.00018476235357479465,
"loss": 2.9355,
"step": 102900
},
{
"epoch": 27.704672695951764,
"grad_norm": 0.4848840534687042,
"learning_rate": 0.00018456038777433685,
"loss": 2.9427,
"step": 102950
},
{
"epoch": 27.718130921619295,
"grad_norm": 0.5028627514839172,
"learning_rate": 0.0001843584219738791,
"loss": 2.9512,
"step": 103000
},
{
"epoch": 27.718130921619295,
"eval_accuracy": 0.3975925160569188,
"eval_loss": 3.2787537574768066,
"eval_runtime": 53.2354,
"eval_samples_per_second": 338.346,
"eval_steps_per_second": 21.151,
"step": 103000
},
{
"epoch": 27.731589147286822,
"grad_norm": 0.5032782554626465,
"learning_rate": 0.0001841564561734213,
"loss": 2.9493,
"step": 103050
},
{
"epoch": 27.74504737295435,
"grad_norm": 0.49334263801574707,
"learning_rate": 0.00018395449037296352,
"loss": 2.9353,
"step": 103100
},
{
"epoch": 27.758505598621877,
"grad_norm": 0.48438459634780884,
"learning_rate": 0.0001837525245725057,
"loss": 2.9498,
"step": 103150
},
{
"epoch": 27.771963824289404,
"grad_norm": 0.5045686364173889,
"learning_rate": 0.0001835505587720479,
"loss": 2.9449,
"step": 103200
},
{
"epoch": 27.785422049956935,
"grad_norm": 0.5195924639701843,
"learning_rate": 0.00018334859297159013,
"loss": 2.9567,
"step": 103250
},
{
"epoch": 27.798880275624462,
"grad_norm": 0.5344165563583374,
"learning_rate": 0.00018314662717113233,
"loss": 2.9542,
"step": 103300
},
{
"epoch": 27.81233850129199,
"grad_norm": 0.4892471730709076,
"learning_rate": 0.00018294466137067455,
"loss": 2.953,
"step": 103350
},
{
"epoch": 27.825796726959517,
"grad_norm": 0.46618539094924927,
"learning_rate": 0.00018274269557021675,
"loss": 2.9543,
"step": 103400
},
{
"epoch": 27.839254952627044,
"grad_norm": 0.49432826042175293,
"learning_rate": 0.000182540729769759,
"loss": 2.9486,
"step": 103450
},
{
"epoch": 27.852713178294575,
"grad_norm": 0.5202397108078003,
"learning_rate": 0.0001823387639693012,
"loss": 2.9531,
"step": 103500
},
{
"epoch": 27.866171403962102,
"grad_norm": 0.5039646625518799,
"learning_rate": 0.0001821367981688434,
"loss": 2.9587,
"step": 103550
},
{
"epoch": 27.87962962962963,
"grad_norm": 0.4717576503753662,
"learning_rate": 0.0001819348323683856,
"loss": 2.9552,
"step": 103600
},
{
"epoch": 27.893087855297157,
"grad_norm": 0.47932833433151245,
"learning_rate": 0.0001817328665679278,
"loss": 2.9671,
"step": 103650
},
{
"epoch": 27.906546080964684,
"grad_norm": 0.47947022318840027,
"learning_rate": 0.00018153090076747003,
"loss": 2.9601,
"step": 103700
},
{
"epoch": 27.920004306632215,
"grad_norm": 0.5247743725776672,
"learning_rate": 0.00018132893496701223,
"loss": 2.959,
"step": 103750
},
{
"epoch": 27.933462532299743,
"grad_norm": 0.4862915873527527,
"learning_rate": 0.00018112696916655442,
"loss": 2.9635,
"step": 103800
},
{
"epoch": 27.94692075796727,
"grad_norm": 0.5073654651641846,
"learning_rate": 0.00018092500336609665,
"loss": 2.9446,
"step": 103850
},
{
"epoch": 27.960378983634797,
"grad_norm": 0.47922730445861816,
"learning_rate": 0.0001807230375656389,
"loss": 2.9513,
"step": 103900
},
{
"epoch": 27.973837209302324,
"grad_norm": 0.48917800188064575,
"learning_rate": 0.0001805210717651811,
"loss": 2.9668,
"step": 103950
},
{
"epoch": 27.987295434969855,
"grad_norm": 0.5510256290435791,
"learning_rate": 0.0001803191059647233,
"loss": 2.9602,
"step": 104000
},
{
"epoch": 27.987295434969855,
"eval_accuracy": 0.39762434958559917,
"eval_loss": 3.276536464691162,
"eval_runtime": 53.2457,
"eval_samples_per_second": 338.281,
"eval_steps_per_second": 21.147,
"step": 104000
},
{
"epoch": 28.0005383290267,
"grad_norm": 0.48279932141304016,
"learning_rate": 0.0001801171401642655,
"loss": 2.9493,
"step": 104050
},
{
"epoch": 28.01399655469423,
"grad_norm": 0.4794139266014099,
"learning_rate": 0.0001799151743638077,
"loss": 2.8847,
"step": 104100
},
{
"epoch": 28.027454780361758,
"grad_norm": 0.5282390117645264,
"learning_rate": 0.00017971320856334993,
"loss": 2.8819,
"step": 104150
},
{
"epoch": 28.040913006029285,
"grad_norm": 0.49238306283950806,
"learning_rate": 0.00017951124276289213,
"loss": 2.8799,
"step": 104200
},
{
"epoch": 28.054371231696813,
"grad_norm": 0.4960058033466339,
"learning_rate": 0.00017930927696243432,
"loss": 2.8872,
"step": 104250
},
{
"epoch": 28.06782945736434,
"grad_norm": 0.5175946950912476,
"learning_rate": 0.00017910731116197655,
"loss": 2.8864,
"step": 104300
},
{
"epoch": 28.08128768303187,
"grad_norm": 0.5203003883361816,
"learning_rate": 0.00017890534536151877,
"loss": 2.8859,
"step": 104350
},
{
"epoch": 28.094745908699398,
"grad_norm": 0.46728551387786865,
"learning_rate": 0.000178703379561061,
"loss": 2.8913,
"step": 104400
},
{
"epoch": 28.108204134366925,
"grad_norm": 0.5124992728233337,
"learning_rate": 0.0001785014137606032,
"loss": 2.8976,
"step": 104450
},
{
"epoch": 28.121662360034453,
"grad_norm": 0.4944259524345398,
"learning_rate": 0.0001782994479601454,
"loss": 2.8971,
"step": 104500
},
{
"epoch": 28.13512058570198,
"grad_norm": 0.500703752040863,
"learning_rate": 0.0001780974821596876,
"loss": 2.8966,
"step": 104550
},
{
"epoch": 28.14857881136951,
"grad_norm": 0.47100183367729187,
"learning_rate": 0.0001778955163592298,
"loss": 2.893,
"step": 104600
},
{
"epoch": 28.162037037037038,
"grad_norm": 0.48921963572502136,
"learning_rate": 0.00017769355055877203,
"loss": 2.8974,
"step": 104650
},
{
"epoch": 28.175495262704565,
"grad_norm": 0.4933052062988281,
"learning_rate": 0.00017749158475831422,
"loss": 2.8969,
"step": 104700
},
{
"epoch": 28.188953488372093,
"grad_norm": 0.5153895616531372,
"learning_rate": 0.00017728961895785645,
"loss": 2.9015,
"step": 104750
},
{
"epoch": 28.20241171403962,
"grad_norm": 0.4876326322555542,
"learning_rate": 0.00017708765315739867,
"loss": 2.9076,
"step": 104800
},
{
"epoch": 28.215869939707147,
"grad_norm": 0.5190147161483765,
"learning_rate": 0.0001768856873569409,
"loss": 2.909,
"step": 104850
},
{
"epoch": 28.229328165374678,
"grad_norm": 0.5012753009796143,
"learning_rate": 0.0001766837215564831,
"loss": 2.9047,
"step": 104900
},
{
"epoch": 28.242786391042205,
"grad_norm": 0.5472180843353271,
"learning_rate": 0.0001764817557560253,
"loss": 2.9069,
"step": 104950
},
{
"epoch": 28.256244616709733,
"grad_norm": 0.5082560181617737,
"learning_rate": 0.0001762797899555675,
"loss": 2.9208,
"step": 105000
},
{
"epoch": 28.256244616709733,
"eval_accuracy": 0.39672627467750354,
"eval_loss": 3.2896358966827393,
"eval_runtime": 53.2461,
"eval_samples_per_second": 338.278,
"eval_steps_per_second": 21.147,
"step": 105000
},
{
"epoch": 28.26970284237726,
"grad_norm": 0.5045067071914673,
"learning_rate": 0.0001760778241551097,
"loss": 2.9129,
"step": 105050
},
{
"epoch": 28.283161068044787,
"grad_norm": 0.5054407119750977,
"learning_rate": 0.00017587585835465193,
"loss": 2.9174,
"step": 105100
},
{
"epoch": 28.296619293712318,
"grad_norm": 0.4975518584251404,
"learning_rate": 0.00017567389255419412,
"loss": 2.9095,
"step": 105150
},
{
"epoch": 28.310077519379846,
"grad_norm": 0.47465768456459045,
"learning_rate": 0.00017547192675373635,
"loss": 2.9034,
"step": 105200
},
{
"epoch": 28.323535745047373,
"grad_norm": 0.5002067685127258,
"learning_rate": 0.00017526996095327857,
"loss": 2.9111,
"step": 105250
},
{
"epoch": 28.3369939707149,
"grad_norm": 0.5059365630149841,
"learning_rate": 0.0001750679951528208,
"loss": 2.9186,
"step": 105300
},
{
"epoch": 28.350452196382427,
"grad_norm": 0.5252241492271423,
"learning_rate": 0.000174866029352363,
"loss": 2.9172,
"step": 105350
},
{
"epoch": 28.36391042204996,
"grad_norm": 0.5054560303688049,
"learning_rate": 0.0001746640635519052,
"loss": 2.9179,
"step": 105400
},
{
"epoch": 28.377368647717486,
"grad_norm": 0.5258662104606628,
"learning_rate": 0.0001744620977514474,
"loss": 2.9187,
"step": 105450
},
{
"epoch": 28.390826873385013,
"grad_norm": 0.4886089861392975,
"learning_rate": 0.0001742601319509896,
"loss": 2.9154,
"step": 105500
},
{
"epoch": 28.40428509905254,
"grad_norm": 0.49281156063079834,
"learning_rate": 0.00017405816615053183,
"loss": 2.9172,
"step": 105550
},
{
"epoch": 28.417743324720067,
"grad_norm": 0.5134946703910828,
"learning_rate": 0.00017385620035007403,
"loss": 2.9255,
"step": 105600
},
{
"epoch": 28.4312015503876,
"grad_norm": 0.532423198223114,
"learning_rate": 0.00017365423454961622,
"loss": 2.9167,
"step": 105650
},
{
"epoch": 28.444659776055126,
"grad_norm": 0.5343100428581238,
"learning_rate": 0.00017345226874915847,
"loss": 2.9168,
"step": 105700
},
{
"epoch": 28.458118001722653,
"grad_norm": 0.5156179070472717,
"learning_rate": 0.00017325030294870067,
"loss": 2.9378,
"step": 105750
},
{
"epoch": 28.47157622739018,
"grad_norm": 0.4972434639930725,
"learning_rate": 0.0001730483371482429,
"loss": 2.918,
"step": 105800
},
{
"epoch": 28.485034453057708,
"grad_norm": 0.4892217218875885,
"learning_rate": 0.0001728463713477851,
"loss": 2.9274,
"step": 105850
},
{
"epoch": 28.49849267872524,
"grad_norm": 0.4946894347667694,
"learning_rate": 0.0001726444055473273,
"loss": 2.9196,
"step": 105900
},
{
"epoch": 28.511950904392766,
"grad_norm": 0.5090630650520325,
"learning_rate": 0.0001724424397468695,
"loss": 2.9302,
"step": 105950
},
{
"epoch": 28.525409130060293,
"grad_norm": 0.5210672616958618,
"learning_rate": 0.00017224047394641173,
"loss": 2.9302,
"step": 106000
},
{
"epoch": 28.525409130060293,
"eval_accuracy": 0.3973465395759209,
"eval_loss": 3.285060405731201,
"eval_runtime": 53.2805,
"eval_samples_per_second": 338.06,
"eval_steps_per_second": 21.133,
"step": 106000
},
{
"epoch": 28.53886735572782,
"grad_norm": 0.516939640045166,
"learning_rate": 0.00017203850814595393,
"loss": 2.9195,
"step": 106050
},
{
"epoch": 28.552325581395348,
"grad_norm": 0.5163007974624634,
"learning_rate": 0.00017183654234549612,
"loss": 2.9206,
"step": 106100
},
{
"epoch": 28.56578380706288,
"grad_norm": 0.5225358009338379,
"learning_rate": 0.00017163457654503837,
"loss": 2.9237,
"step": 106150
},
{
"epoch": 28.579242032730406,
"grad_norm": 0.5029885172843933,
"learning_rate": 0.00017143261074458057,
"loss": 2.9381,
"step": 106200
},
{
"epoch": 28.592700258397933,
"grad_norm": 0.4904434084892273,
"learning_rate": 0.0001712306449441228,
"loss": 2.9269,
"step": 106250
},
{
"epoch": 28.60615848406546,
"grad_norm": 0.49272313714027405,
"learning_rate": 0.000171028679143665,
"loss": 2.9339,
"step": 106300
},
{
"epoch": 28.619616709732988,
"grad_norm": 0.5093855261802673,
"learning_rate": 0.0001708267133432072,
"loss": 2.935,
"step": 106350
},
{
"epoch": 28.63307493540052,
"grad_norm": 0.4888289272785187,
"learning_rate": 0.0001706247475427494,
"loss": 2.9396,
"step": 106400
},
{
"epoch": 28.646533161068046,
"grad_norm": 0.4979041814804077,
"learning_rate": 0.0001704227817422916,
"loss": 2.9331,
"step": 106450
},
{
"epoch": 28.659991386735573,
"grad_norm": 0.5191659927368164,
"learning_rate": 0.00017022081594183383,
"loss": 2.9327,
"step": 106500
},
{
"epoch": 28.6734496124031,
"grad_norm": 0.48608386516571045,
"learning_rate": 0.00017001885014137605,
"loss": 2.9355,
"step": 106550
},
{
"epoch": 28.686907838070628,
"grad_norm": 0.47546452283859253,
"learning_rate": 0.00016981688434091827,
"loss": 2.9403,
"step": 106600
},
{
"epoch": 28.70036606373816,
"grad_norm": 0.5148166418075562,
"learning_rate": 0.00016961491854046047,
"loss": 2.9391,
"step": 106650
},
{
"epoch": 28.713824289405686,
"grad_norm": 0.4863901734352112,
"learning_rate": 0.0001694129527400027,
"loss": 2.9444,
"step": 106700
},
{
"epoch": 28.727282515073213,
"grad_norm": 0.5058971643447876,
"learning_rate": 0.0001692109869395449,
"loss": 2.9366,
"step": 106750
},
{
"epoch": 28.74074074074074,
"grad_norm": 0.511207640171051,
"learning_rate": 0.00016900902113908708,
"loss": 2.9372,
"step": 106800
},
{
"epoch": 28.754198966408268,
"grad_norm": 0.5124873518943787,
"learning_rate": 0.0001688070553386293,
"loss": 2.9434,
"step": 106850
},
{
"epoch": 28.767657192075795,
"grad_norm": 0.4950677454471588,
"learning_rate": 0.0001686050895381715,
"loss": 2.9411,
"step": 106900
},
{
"epoch": 28.781115417743326,
"grad_norm": 0.5125149488449097,
"learning_rate": 0.00016840312373771373,
"loss": 2.9478,
"step": 106950
},
{
"epoch": 28.794573643410853,
"grad_norm": 0.5346261858940125,
"learning_rate": 0.00016820115793725595,
"loss": 2.937,
"step": 107000
},
{
"epoch": 28.794573643410853,
"eval_accuracy": 0.39771050654206175,
"eval_loss": 3.278653144836426,
"eval_runtime": 53.1172,
"eval_samples_per_second": 339.099,
"eval_steps_per_second": 21.198,
"step": 107000
},
{
"epoch": 28.80803186907838,
"grad_norm": 0.48506495356559753,
"learning_rate": 0.00016799919213679817,
"loss": 2.9377,
"step": 107050
},
{
"epoch": 28.821490094745908,
"grad_norm": 0.48778578639030457,
"learning_rate": 0.00016779722633634037,
"loss": 2.9432,
"step": 107100
},
{
"epoch": 28.834948320413435,
"grad_norm": 0.501910924911499,
"learning_rate": 0.0001675952605358826,
"loss": 2.9345,
"step": 107150
},
{
"epoch": 28.848406546080966,
"grad_norm": 0.48859545588493347,
"learning_rate": 0.0001673932947354248,
"loss": 2.9464,
"step": 107200
},
{
"epoch": 28.861864771748493,
"grad_norm": 0.5098825693130493,
"learning_rate": 0.00016719132893496698,
"loss": 2.9489,
"step": 107250
},
{
"epoch": 28.87532299741602,
"grad_norm": 0.49968665838241577,
"learning_rate": 0.0001669893631345092,
"loss": 2.937,
"step": 107300
},
{
"epoch": 28.888781223083548,
"grad_norm": 0.5197726488113403,
"learning_rate": 0.0001667873973340514,
"loss": 2.9373,
"step": 107350
},
{
"epoch": 28.902239448751075,
"grad_norm": 0.5304186344146729,
"learning_rate": 0.00016658543153359363,
"loss": 2.9492,
"step": 107400
},
{
"epoch": 28.915697674418606,
"grad_norm": 0.48765209317207336,
"learning_rate": 0.00016638346573313585,
"loss": 2.9455,
"step": 107450
},
{
"epoch": 28.929155900086133,
"grad_norm": 0.48178255558013916,
"learning_rate": 0.00016618149993267807,
"loss": 2.9384,
"step": 107500
},
{
"epoch": 28.94261412575366,
"grad_norm": 0.5031225085258484,
"learning_rate": 0.00016597953413222027,
"loss": 2.9479,
"step": 107550
},
{
"epoch": 28.956072351421188,
"grad_norm": 0.5146300196647644,
"learning_rate": 0.00016577756833176247,
"loss": 2.9304,
"step": 107600
},
{
"epoch": 28.969530577088715,
"grad_norm": 0.5076146125793457,
"learning_rate": 0.0001655756025313047,
"loss": 2.9531,
"step": 107650
},
{
"epoch": 28.982988802756246,
"grad_norm": 0.5111352205276489,
"learning_rate": 0.00016537363673084688,
"loss": 2.9426,
"step": 107700
},
{
"epoch": 28.996447028423773,
"grad_norm": 0.5215750336647034,
"learning_rate": 0.0001651716709303891,
"loss": 2.9438,
"step": 107750
},
{
"epoch": 29.00968992248062,
"grad_norm": 0.5104740262031555,
"learning_rate": 0.0001649697051299313,
"loss": 2.8855,
"step": 107800
},
{
"epoch": 29.02314814814815,
"grad_norm": 0.5092785954475403,
"learning_rate": 0.0001647677393294735,
"loss": 2.8753,
"step": 107850
},
{
"epoch": 29.036606373815676,
"grad_norm": 0.4839480221271515,
"learning_rate": 0.00016456577352901575,
"loss": 2.8683,
"step": 107900
},
{
"epoch": 29.050064599483203,
"grad_norm": 0.5309181809425354,
"learning_rate": 0.00016436380772855797,
"loss": 2.8825,
"step": 107950
},
{
"epoch": 29.06352282515073,
"grad_norm": 0.5218488574028015,
"learning_rate": 0.00016416184192810017,
"loss": 2.877,
"step": 108000
},
{
"epoch": 29.06352282515073,
"eval_accuracy": 0.3972138817652767,
"eval_loss": 3.2904505729675293,
"eval_runtime": 53.29,
"eval_samples_per_second": 338.0,
"eval_steps_per_second": 21.13,
"step": 108000
},
{
"epoch": 29.07698105081826,
"grad_norm": 0.5029823184013367,
"learning_rate": 0.00016395987612764237,
"loss": 2.8809,
"step": 108050
},
{
"epoch": 29.09043927648579,
"grad_norm": 0.5229007005691528,
"learning_rate": 0.0001637579103271846,
"loss": 2.8787,
"step": 108100
},
{
"epoch": 29.103897502153316,
"grad_norm": 0.5105121731758118,
"learning_rate": 0.00016355594452672679,
"loss": 2.8975,
"step": 108150
},
{
"epoch": 29.117355727820843,
"grad_norm": 0.5129608511924744,
"learning_rate": 0.000163353978726269,
"loss": 2.8836,
"step": 108200
},
{
"epoch": 29.13081395348837,
"grad_norm": 0.5075309872627258,
"learning_rate": 0.0001631520129258112,
"loss": 2.8903,
"step": 108250
},
{
"epoch": 29.1442721791559,
"grad_norm": 0.5150623321533203,
"learning_rate": 0.0001629500471253534,
"loss": 2.8957,
"step": 108300
},
{
"epoch": 29.15773040482343,
"grad_norm": 0.5157734155654907,
"learning_rate": 0.00016274808132489565,
"loss": 2.895,
"step": 108350
},
{
"epoch": 29.171188630490956,
"grad_norm": 0.5129877328872681,
"learning_rate": 0.00016254611552443785,
"loss": 2.8948,
"step": 108400
},
{
"epoch": 29.184646856158484,
"grad_norm": 0.49951955676078796,
"learning_rate": 0.00016234414972398007,
"loss": 2.8912,
"step": 108450
},
{
"epoch": 29.19810508182601,
"grad_norm": 0.5277801156044006,
"learning_rate": 0.00016214218392352227,
"loss": 2.8945,
"step": 108500
},
{
"epoch": 29.21156330749354,
"grad_norm": 0.5068605542182922,
"learning_rate": 0.0001619402181230645,
"loss": 2.8926,
"step": 108550
},
{
"epoch": 29.22502153316107,
"grad_norm": 0.5398867130279541,
"learning_rate": 0.00016173825232260669,
"loss": 2.9044,
"step": 108600
},
{
"epoch": 29.238479758828596,
"grad_norm": 0.49381494522094727,
"learning_rate": 0.00016153628652214888,
"loss": 2.9003,
"step": 108650
},
{
"epoch": 29.251937984496124,
"grad_norm": 0.5253661274909973,
"learning_rate": 0.0001613343207216911,
"loss": 2.9056,
"step": 108700
},
{
"epoch": 29.26539621016365,
"grad_norm": 0.516562819480896,
"learning_rate": 0.0001611323549212333,
"loss": 2.906,
"step": 108750
},
{
"epoch": 29.27885443583118,
"grad_norm": 0.5054872632026672,
"learning_rate": 0.00016093038912077555,
"loss": 2.9018,
"step": 108800
},
{
"epoch": 29.29231266149871,
"grad_norm": 0.5125064849853516,
"learning_rate": 0.00016072842332031775,
"loss": 2.91,
"step": 108850
},
{
"epoch": 29.305770887166236,
"grad_norm": 0.5209744572639465,
"learning_rate": 0.00016052645751985997,
"loss": 2.8966,
"step": 108900
},
{
"epoch": 29.319229112833764,
"grad_norm": 0.5007820725440979,
"learning_rate": 0.00016032449171940217,
"loss": 2.9104,
"step": 108950
},
{
"epoch": 29.33268733850129,
"grad_norm": 0.523684024810791,
"learning_rate": 0.0001601225259189444,
"loss": 2.9082,
"step": 109000
},
{
"epoch": 29.33268733850129,
"eval_accuracy": 0.39706155887377537,
"eval_loss": 3.288200855255127,
"eval_runtime": 53.168,
"eval_samples_per_second": 338.775,
"eval_steps_per_second": 21.178,
"step": 109000
},
{
"epoch": 29.33268733850129,
"step": 109000,
"total_flos": 2.277981561028608e+18,
"train_loss": 3.1618109384239266,
"train_runtime": 42692.4106,
"train_samples_per_second": 278.472,
"train_steps_per_second": 3.482
}
],
"logging_steps": 50,
"max_steps": 148640,
"num_input_tokens_seen": 0,
"num_train_epochs": 40,
"save_steps": 10000,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 20,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 11
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.277981561028608e+18,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}
{
"best_global_step": 89000,
"best_metric": 3.275252103805542,
"best_model_checkpoint": "/scratch/cl5625/exceptions/models/high_2000_6910/checkpoint-70000",
"epoch": 29.33268733850129,
"eval_steps": 1000,
"global_step": 109000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.013458225667527994,
"grad_norm": 3.308773994445801,
"learning_rate": 0.000294,
"loss": 8.4679,
"step": 50
},
{
"epoch": 0.026916451335055987,
"grad_norm": 1.6042652130126953,
"learning_rate": 0.0005939999999999999,
"loss": 6.7351,
"step": 100
},
{
"epoch": 0.04037467700258398,
"grad_norm": 1.6990554332733154,
"learning_rate": 0.0005998020735155513,
"loss": 6.3185,
"step": 150
},
{
"epoch": 0.053832902670111975,
"grad_norm": 0.9088912010192871,
"learning_rate": 0.0005996001077150935,
"loss": 6.0464,
"step": 200
},
{
"epoch": 0.06729112833763996,
"grad_norm": 0.6608636379241943,
"learning_rate": 0.0005993981419146358,
"loss": 5.8696,
"step": 250
},
{
"epoch": 0.08074935400516796,
"grad_norm": 1.4318311214447021,
"learning_rate": 0.0005991961761141779,
"loss": 5.7427,
"step": 300
},
{
"epoch": 0.09420757967269595,
"grad_norm": 2.375964641571045,
"learning_rate": 0.0005989942103137202,
"loss": 5.6263,
"step": 350
},
{
"epoch": 0.10766580534022395,
"grad_norm": 1.0631674528121948,
"learning_rate": 0.0005987922445132624,
"loss": 5.4973,
"step": 400
},
{
"epoch": 0.12112403100775193,
"grad_norm": 1.3984321355819702,
"learning_rate": 0.0005985902787128047,
"loss": 5.3815,
"step": 450
},
{
"epoch": 0.13458225667527993,
"grad_norm": 1.0844402313232422,
"learning_rate": 0.0005983883129123468,
"loss": 5.2436,
"step": 500
},
{
"epoch": 0.14804048234280792,
"grad_norm": 1.4188182353973389,
"learning_rate": 0.000598186347111889,
"loss": 5.1461,
"step": 550
},
{
"epoch": 0.16149870801033592,
"grad_norm": 1.540555715560913,
"learning_rate": 0.0005979843813114312,
"loss": 5.0629,
"step": 600
},
{
"epoch": 0.1749569336778639,
"grad_norm": 1.31498122215271,
"learning_rate": 0.0005977824155109734,
"loss": 4.9972,