Text Generation
Transformers
Safetensors
English
bananamind2_pico
causal-lm
base-model
muon
xsa-refresh
custom-code
trust-remote-code
custom_code
Instructions to use Banaxi-Tech/pico-test with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Banaxi-Tech/pico-test with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Banaxi-Tech/pico-test", trust_remote_code=True)# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("Banaxi-Tech/pico-test", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use Banaxi-Tech/pico-test with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Banaxi-Tech/pico-test" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Banaxi-Tech/pico-test", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/Banaxi-Tech/pico-test
- SGLang
How to use Banaxi-Tech/pico-test 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 "Banaxi-Tech/pico-test" \ --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": "Banaxi-Tech/pico-test", "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 "Banaxi-Tech/pico-test" \ --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": "Banaxi-Tech/pico-test", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use Banaxi-Tech/pico-test with Docker Model Runner:
docker model run hf.co/Banaxi-Tech/pico-test
Upload BananaMind 2 Pico 90% checkpoint at 180,004,847,616 tokens
Browse files- README.md +4 -4
- checkpoint_metadata.json +6 -6
- model.safetensors +1 -1
- training_metrics.jsonl +112 -0
README.md
CHANGED
|
@@ -15,9 +15,9 @@ tags:
|
|
| 15 |
- trust-remote-code
|
| 16 |
---
|
| 17 |
|
| 18 |
-
# BananaMind 2 Pico Test -
|
| 19 |
|
| 20 |
-
This is the **
|
| 21 |
base causal language model. It is not instruction tuned.
|
| 22 |
|
| 23 |
## Architecture
|
|
@@ -42,8 +42,8 @@ history alongside the K/V cache. Its learned residual scalar starts at zero.
|
|
| 42 |
|
| 43 |
| Field | Value |
|
| 44 |
|---|---:|
|
| 45 |
-
| Progress |
|
| 46 |
-
| Tokens seen |
|
| 47 |
| Target tokens | 200,000,000,000 |
|
| 48 |
| Hardware | 4 x NVIDIA H200 |
|
| 49 |
| Matrix optimizer | Stock `torch.optim.Muon` |
|
|
|
|
| 15 |
- trust-remote-code
|
| 16 |
---
|
| 17 |
|
| 18 |
+
# BananaMind 2 Pico Test - 90%
|
| 19 |
|
| 20 |
+
This is the **90% checkpoint** of a 900,002-parameter
|
| 21 |
base causal language model. It is not instruction tuned.
|
| 22 |
|
| 23 |
## Architecture
|
|
|
|
| 42 |
|
| 43 |
| Field | Value |
|
| 44 |
|---|---:|
|
| 45 |
+
| Progress | 90% |
|
| 46 |
+
| Tokens seen | 180,004,847,616 |
|
| 47 |
| Target tokens | 200,000,000,000 |
|
| 48 |
| Hardware | 4 x NVIDIA H200 |
|
| 49 |
| Matrix optimizer | Stock `torch.optim.Muon` |
|
checkpoint_metadata.json
CHANGED
|
@@ -44,10 +44,10 @@
|
|
| 44 |
"model_type": "bananamind2_pico",
|
| 45 |
"output_attentions": false
|
| 46 |
},
|
| 47 |
-
"training_percent":
|
| 48 |
-
"step":
|
| 49 |
"total_steps": 22439,
|
| 50 |
-
"tokens_seen":
|
| 51 |
"target_tokens": 200000000000,
|
| 52 |
"trained_token_target": 199996473344,
|
| 53 |
"tokens_per_step": 8912896,
|
|
@@ -65,8 +65,8 @@
|
|
| 65 |
"cosmopedia_v2": "3ba9d605774198c5868892d7a8deda78031a781f"
|
| 66 |
},
|
| 67 |
"source_tokens": {
|
| 68 |
-
"fineweb_hq":
|
| 69 |
-
"cosmopedia_v2":
|
| 70 |
},
|
| 71 |
-
"elapsed_seconds":
|
| 72 |
}
|
|
|
|
| 44 |
"model_type": "bananamind2_pico",
|
| 45 |
"output_attentions": false
|
| 46 |
},
|
| 47 |
+
"training_percent": 90,
|
| 48 |
+
"step": 20196,
|
| 49 |
"total_steps": 22439,
|
| 50 |
+
"tokens_seen": 180004847616,
|
| 51 |
"target_tokens": 200000000000,
|
| 52 |
"trained_token_target": 199996473344,
|
| 53 |
"tokens_per_step": 8912896,
|
|
|
|
| 65 |
"cosmopedia_v2": "3ba9d605774198c5868892d7a8deda78031a781f"
|
| 66 |
},
|
| 67 |
"source_tokens": {
|
| 68 |
+
"fineweb_hq": 128007012352,
|
| 69 |
+
"cosmopedia_v2": 51997835264
|
| 70 |
},
|
| 71 |
+
"elapsed_seconds": 23685.649554491043
|
| 72 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3756344
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7597cbfd8e45f59e61727a44a7129fbc10c246c7635123265b61a64cc736fe06
|
| 3 |
size 3756344
|
training_metrics.jsonl
CHANGED
|
@@ -1906,3 +1906,115 @@
|
|
| 1906 |
{"step": 19050, "total_steps": 22439, "tokens": 169790668800, "source": "fineweb_hq", "phase": 2, "loss": 1.8293047547340393, "ce_loss": 1.8293047547340393, "logit_z_loss": 101.95938301086426, "muon_lr": 0.07, "adamw_lr": 0.004, "grad_norm": 0.3222925662994385, "tokens_per_second": 9051288.675644593, "eta_seconds": 3337.18276219368, "data_wait_seconds": 1.9669532775878906e-05}
|
| 1907 |
{"step": 19060, "total_steps": 22439, "tokens": 169879797760, "source": "fineweb_hq", "phase": 2, "loss": 1.814415603876114, "ce_loss": 1.814415603876114, "logit_z_loss": 106.60835647583008, "muon_lr": 0.07, "adamw_lr": 0.004, "grad_norm": 0.3672824501991272, "tokens_per_second": 9649960.09089034, "eta_seconds": 3120.9119312763214, "data_wait_seconds": 3.11732292175293e-05}
|
| 1908 |
{"step": 19070, "total_steps": 22439, "tokens": 169968926720, "source": "fineweb_hq", "phase": 2, "loss": 1.8197243809700012, "ce_loss": 1.8197243809700012, "logit_z_loss": 109.8288345336914, "muon_lr": 0.07, "adamw_lr": 0.004, "grad_norm": 0.33823081851005554, "tokens_per_second": 10621130.950083543, "eta_seconds": 2827.151530766487, "data_wait_seconds": 1.6868114471435547e-05}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1906 |
{"step": 19050, "total_steps": 22439, "tokens": 169790668800, "source": "fineweb_hq", "phase": 2, "loss": 1.8293047547340393, "ce_loss": 1.8293047547340393, "logit_z_loss": 101.95938301086426, "muon_lr": 0.07, "adamw_lr": 0.004, "grad_norm": 0.3222925662994385, "tokens_per_second": 9051288.675644593, "eta_seconds": 3337.18276219368, "data_wait_seconds": 1.9669532775878906e-05}
|
| 1907 |
{"step": 19060, "total_steps": 22439, "tokens": 169879797760, "source": "fineweb_hq", "phase": 2, "loss": 1.814415603876114, "ce_loss": 1.814415603876114, "logit_z_loss": 106.60835647583008, "muon_lr": 0.07, "adamw_lr": 0.004, "grad_norm": 0.3672824501991272, "tokens_per_second": 9649960.09089034, "eta_seconds": 3120.9119312763214, "data_wait_seconds": 3.11732292175293e-05}
|
| 1908 |
{"step": 19070, "total_steps": 22439, "tokens": 169968926720, "source": "fineweb_hq", "phase": 2, "loss": 1.8197243809700012, "ce_loss": 1.8197243809700012, "logit_z_loss": 109.8288345336914, "muon_lr": 0.07, "adamw_lr": 0.004, "grad_norm": 0.33823081851005554, "tokens_per_second": 10621130.950083543, "eta_seconds": 2827.151530766487, "data_wait_seconds": 1.6868114471435547e-05}
|
| 1909 |
+
{"step": 19080, "total_steps": 22439, "tokens": 170058055680, "source": "fineweb_hq", "phase": 2, "loss": 1.8115954399108887, "ce_loss": 1.8115954399108887, "logit_z_loss": 106.31680679321289, "muon_lr": 0.06999961843806778, "adamw_lr": 0.003999978196461015, "grad_norm": 0.3432227075099945, "tokens_per_second": 10445125.141346756, "eta_seconds": 2866.257441520691, "data_wait_seconds": 1.615285873413086e-05}
|
| 1910 |
+
{"step": 19090, "total_steps": 22439, "tokens": 170147184640, "source": "fineweb_hq", "phase": 2, "loss": 1.8042868077754974, "ce_loss": 1.8042868077754974, "logit_z_loss": 104.24834632873535, "muon_lr": 0.0699965659925261, "adamw_lr": 0.003999803771001491, "grad_norm": 0.3229518532752991, "tokens_per_second": 9998269.141846372, "eta_seconds": 2985.445608687401, "data_wait_seconds": 3.045797348022461e-05}
|
| 1911 |
+
{"step": 19100, "total_steps": 22439, "tokens": 170236313600, "source": "fineweb_hq", "phase": 2, "loss": 1.8192864060401917, "ce_loss": 1.8192864060401917, "logit_z_loss": 106.19583320617676, "muon_lr": 0.06999046136765774, "adamw_lr": 0.003999454935294728, "grad_norm": 0.3156251609325409, "tokens_per_second": 9834818.816730933, "eta_seconds": 3025.999797105789, "data_wait_seconds": 2.086162567138672e-05}
|
| 1912 |
+
{"step": 19110, "total_steps": 22439, "tokens": 170325442560, "source": "fineweb_hq", "phase": 2, "loss": 1.790729284286499, "ce_loss": 1.790729284286499, "logit_z_loss": 107.38431549072266, "muon_lr": 0.06998130509586954, "adamw_lr": 0.003998931719763974, "grad_norm": 0.3806309401988983, "tokens_per_second": 8863088.58748263, "eta_seconds": 3347.707798600197, "data_wait_seconds": 2.0265579223632812e-05}
|
| 1913 |
+
{"step": 19120, "total_steps": 22439, "tokens": 170414571520, "source": "fineweb_hq", "phase": 2, "loss": 1.8239244520664215, "ce_loss": 1.8239244520664215, "logit_z_loss": 106.17072868347168, "muon_lr": 0.06996909797571364, "adamw_lr": 0.003998234170040779, "grad_norm": 0.32472944259643555, "tokens_per_second": 9216156.546727402, "eta_seconds": 3209.78725502491, "data_wait_seconds": 1.7583370208740234e-05}
|
| 1914 |
+
{"step": 19130, "total_steps": 22439, "tokens": 170503700480, "source": "fineweb_hq", "phase": 2, "loss": 1.8479870855808258, "ce_loss": 1.8479870855808258, "logit_z_loss": 103.2541275024414, "muon_lr": 0.0699538410718179, "adamw_lr": 0.003997362346961023, "grad_norm": 0.36446040868759155, "tokens_per_second": 11197058.5214468, "eta_seconds": 2633.9750576019287, "data_wait_seconds": 1.537799835205078e-05}
|
| 1915 |
+
{"step": 19140, "total_steps": 22439, "tokens": 170592829440, "source": "fineweb_hq", "phase": 2, "loss": 1.8761783838272095, "ce_loss": 1.8761783838272095, "logit_z_loss": 106.23115158081055, "muon_lr": 0.0699355357147931, "adamw_lr": 0.0039963163265596055, "grad_norm": 0.2976381480693817, "tokens_per_second": 8193222.774113968, "eta_seconds": 3588.776323390007, "data_wait_seconds": 1.5974044799804688e-05}
|
| 1916 |
+
{"step": 19150, "total_steps": 22439, "tokens": 170681958400, "source": "fineweb_hq", "phase": 2, "loss": 1.9856676757335663, "ce_loss": 1.9856676757335663, "logit_z_loss": 100.51487159729004, "muon_lr": 0.06991418350111676, "adamw_lr": 0.0039950962000638145, "grad_norm": 0.3313066065311432, "tokens_per_second": 9567000.088491576, "eta_seconds": 3064.128219175339, "data_wait_seconds": 1.6510486602783203e-05}
|
| 1917 |
+
{"step": 19160, "total_steps": 22439, "tokens": 170771087360, "source": "fineweb_hq", "phase": 2, "loss": 1.980131834745407, "ce_loss": 1.980131834745407, "logit_z_loss": 97.70445823669434, "muon_lr": 0.06988978629299405, "adamw_lr": 0.003993702073885374, "grad_norm": 0.2938137352466583, "tokens_per_second": 8268835.493693119, "eta_seconds": 3534.401670742035, "data_wait_seconds": 2.104043960571289e-05}
|
| 1918 |
+
{"step": 19170, "total_steps": 22439, "tokens": 170860216320, "source": "fineweb_hq", "phase": 2, "loss": 1.9776781499385834, "ce_loss": 1.9776781499385834, "logit_z_loss": 98.32389450073242, "muon_lr": 0.06986234621819529, "adamw_lr": 0.0039921340696111585, "grad_norm": 0.349571168422699, "tokens_per_second": 9974126.281830028, "eta_seconds": 2921.1838912725448, "data_wait_seconds": 2.104043960571289e-05}
|
| 1919 |
+
{"step": 19180, "total_steps": 22439, "tokens": 170949345280, "source": "fineweb_hq", "phase": 2, "loss": 1.9868899285793304, "ce_loss": 1.9868899285793304, "logit_z_loss": 95.63092613220215, "muon_lr": 0.06983186566987037, "adamw_lr": 0.003990392323992592, "grad_norm": 0.30252572894096375, "tokens_per_second": 10277359.966886703, "eta_seconds": 2826.3219501495364, "data_wait_seconds": 1.710653305053711e-05}
|
| 1920 |
+
{"step": 19190, "total_steps": 22439, "tokens": 171038474240, "source": "fineweb_hq", "phase": 2, "loss": 1.9904397130012512, "ce_loss": 1.9904397130012512, "logit_z_loss": 93.9439525604248, "muon_lr": 0.0697983473063401, "adamw_lr": 0.00398847698893372, "grad_norm": 0.36928847432136536, "tokens_per_second": 8323464.758445735, "eta_seconds": 3479.0799197673796, "data_wait_seconds": 1.8298625946044922e-05}
|
| 1921 |
+
{"step": 19200, "total_steps": 22439, "tokens": 171127603200, "source": "fineweb_hq", "phase": 2, "loss": 1.981736421585083, "ce_loss": 1.981736421585083, "logit_z_loss": 99.16025352478027, "muon_lr": 0.06976179405086437, "adamw_lr": 0.003986388231477963, "grad_norm": 0.32141637802124023, "tokens_per_second": 9936173.813873721, "eta_seconds": 2905.43127412796, "data_wait_seconds": 1.5556812286376953e-05}
|
| 1922 |
+
{"step": 19210, "total_steps": 22439, "tokens": 171216732160, "source": "fineweb_hq", "phase": 2, "loss": 1.9831019043922424, "ce_loss": 1.9831019043922424, "logit_z_loss": 99.28917503356934, "muon_lr": 0.06972220909138707, "adamw_lr": 0.003984126233793546, "grad_norm": 0.32245954871177673, "tokens_per_second": 9101676.787661877, "eta_seconds": 3162.0262788295745, "data_wait_seconds": 2.1636486053466797e-05}
|
| 1923 |
+
{"step": 19220, "total_steps": 22439, "tokens": 171305861120, "source": "fineweb_hq", "phase": 2, "loss": 2.010662853717804, "ce_loss": 2.010662853717804, "logit_z_loss": 94.84764862060547, "muon_lr": 0.06967959588025824, "adamw_lr": 0.003981691193157614, "grad_norm": 0.3678632378578186, "tokens_per_second": 8592753.694418402, "eta_seconds": 3338.931062650681, "data_wait_seconds": 1.5079975128173828e-05}
|
| 1924 |
+
{"step": 19230, "total_steps": 22439, "tokens": 171394990080, "source": "fineweb_hq", "phase": 2, "loss": 1.991280049085617, "ce_loss": 1.991280049085617, "logit_z_loss": 98.23171043395996, "muon_lr": 0.06963395813393286, "adamw_lr": 0.00397908332193902, "grad_norm": 0.31219643354415894, "tokens_per_second": 9025223.48481673, "eta_seconds": 3169.0609448194505, "data_wait_seconds": 1.71661376953125e-05}
|
| 1925 |
+
{"step": 19240, "total_steps": 22439, "tokens": 171484119040, "source": "fineweb_hq", "phase": 2, "loss": 1.9902754127979279, "ce_loss": 1.9902754127979279, "logit_z_loss": 94.4031810760498, "muon_lr": 0.06958529983264676, "adamw_lr": 0.003976302847579815, "grad_norm": 0.3126307725906372, "tokens_per_second": 10346231.37999964, "eta_seconds": 2755.8202843904496, "data_wait_seconds": 1.9669532775878906e-05}
|
| 1926 |
+
{"step": 19250, "total_steps": 22439, "tokens": 171573248000, "source": "fineweb_hq", "phase": 2, "loss": 1.9907587468624115, "ce_loss": 1.9907587468624115, "logit_z_loss": 97.6782283782959, "muon_lr": 0.06953362522006948, "adamw_lr": 0.003973350012575399, "grad_norm": 0.3245856463909149, "tokens_per_second": 8155450.351013346, "eta_seconds": 3485.1815804958346, "data_wait_seconds": 1.913309097290039e-05}
|
| 1927 |
+
{"step": 19260, "total_steps": 22439, "tokens": 171662376960, "source": "fineweb_hq", "phase": 2, "loss": 2.0038043558597565, "ce_loss": 2.0038043558597565, "logit_z_loss": 96.15466690063477, "muon_lr": 0.06947893880293417, "adamw_lr": 0.003970225074453381, "grad_norm": 0.3840566873550415, "tokens_per_second": 9899792.99034487, "eta_seconds": 2862.0897842645645, "data_wait_seconds": 1.6748905181884766e-05}
|
| 1928 |
+
{"step": 19270, "total_steps": 22439, "tokens": 171751505920, "source": "fineweb_hq", "phase": 2, "loss": 2.002715051174164, "ce_loss": 2.002715051174164, "logit_z_loss": 97.47039413452148, "muon_lr": 0.06942124535064453, "adamw_lr": 0.003966928305751116, "grad_norm": 0.331316202878952, "tokens_per_second": 8771835.491537767, "eta_seconds": 3219.9609136819836, "data_wait_seconds": 1.5556812286376953e-05}
|
| 1929 |
+
{"step": 19280, "total_steps": 22439, "tokens": 171840634880, "source": "fineweb_hq", "phase": 2, "loss": 2.007376551628113, "ce_loss": 2.007376551628113, "logit_z_loss": 92.2960205078125, "muon_lr": 0.06936054989485893, "adamw_lr": 0.003963459993991938, "grad_norm": 0.34761422872543335, "tokens_per_second": 8102116.613455535, "eta_seconds": 3475.1213549852373, "data_wait_seconds": 1.7642974853515625e-05}
|
| 1930 |
+
{"step": 19290, "total_steps": 22439, "tokens": 171929763840, "source": "fineweb_hq", "phase": 2, "loss": 1.9922126531600952, "ce_loss": 1.9922126531600952, "logit_z_loss": 95.15262794494629, "muon_lr": 0.0692968577290514, "adamw_lr": 0.00395982044166008, "grad_norm": 0.33466118574142456, "tokens_per_second": 10150859.996643297, "eta_seconds": 2764.9587831258773, "data_wait_seconds": 1.5676021575927734e-05}
|
| 1931 |
+
{"step": 19300, "total_steps": 22439, "tokens": 172018892800, "source": "fineweb_hq", "phase": 2, "loss": 1.956406444311142, "ce_loss": 1.956406444311142, "logit_z_loss": 102.03507232666016, "muon_lr": 0.0692301744080502, "adamw_lr": 0.003956009966174297, "grad_norm": 0.29166167974472046, "tokens_per_second": 7579696.697412474, "eta_seconds": 3691.1213813543322, "data_wait_seconds": 1.7821788787841797e-05}
|
| 1932 |
+
{"step": 19310, "total_steps": 22439, "tokens": 172108021760, "source": "fineweb_hq", "phase": 2, "loss": 1.9349212050437927, "ce_loss": 1.9349212050437927, "logit_z_loss": 98.30118179321289, "muon_lr": 0.06916050574755313, "adamw_lr": 0.003952028899860178, "grad_norm": 0.3933544456958771, "tokens_per_second": 8490961.62544855, "eta_seconds": 3284.486824250221, "data_wait_seconds": 1.9311904907226562e-05}
|
| 1933 |
+
{"step": 19320, "total_steps": 22439, "tokens": 172197150720, "source": "fineweb_hq", "phase": 2, "loss": 1.9356444478034973, "ce_loss": 1.9356444478034973, "logit_z_loss": 101.07486343383789, "muon_lr": 0.06908785782362052, "adamw_lr": 0.0039478775899211725, "grad_norm": 0.32795894145965576, "tokens_per_second": 10006939.534098547, "eta_seconds": 2778.0044567346577, "data_wait_seconds": 1.7404556274414062e-05}
|
| 1934 |
+
{"step": 19330, "total_steps": 22439, "tokens": 172286279680, "source": "fineweb_hq", "phase": 2, "loss": 1.9443558156490326, "ce_loss": 1.9443558156490326, "logit_z_loss": 100.4367790222168, "muon_lr": 0.0690122369721452, "adamw_lr": 0.003943556398408297, "grad_norm": 0.404300332069397, "tokens_per_second": 8138365.985517549, "eta_seconds": 3404.884188461304, "data_wait_seconds": 1.9431114196777344e-05}
|
| 1935 |
+
{"step": 19340, "total_steps": 22439, "tokens": 172375408640, "source": "fineweb_hq", "phase": 2, "loss": 1.9220384061336517, "ce_loss": 1.9220384061336517, "logit_z_loss": 96.64208984375, "muon_lr": 0.06893364978829995, "adamw_lr": 0.003939065702188568, "grad_norm": 0.3583891987800598, "tokens_per_second": 10645137.34948992, "eta_seconds": 2594.711913728714, "data_wait_seconds": 2.014636993408203e-05}
|
| 1936 |
+
{"step": 19350, "total_steps": 22439, "tokens": 172464537600, "source": "fineweb_hq", "phase": 2, "loss": 1.9365795254707336, "ce_loss": 1.9365795254707336, "logit_z_loss": 103.13103103637695, "muon_lr": 0.06885210312596231, "adamw_lr": 0.003934405892912132, "grad_norm": 0.3244110345840454, "tokens_per_second": 8403431.135333734, "eta_seconds": 3276.2731437444686, "data_wait_seconds": 1.913309097290039e-05}
|
| 1937 |
+
{"step": 19360, "total_steps": 22439, "tokens": 172553666560, "source": "fineweb_hq", "phase": 2, "loss": 1.9244379699230194, "ce_loss": 1.9244379699230194, "logit_z_loss": 100.62914657592773, "muon_lr": 0.06876760409711685, "adamw_lr": 0.003929577376978105, "grad_norm": 0.3278604745864868, "tokens_per_second": 10570189.080732115, "eta_seconds": 2596.2455897808077, "data_wait_seconds": 1.8596649169921875e-05}
|
| 1938 |
+
{"step": 19370, "total_steps": 22439, "tokens": 172642795520, "source": "fineweb_hq", "phase": 2, "loss": 1.9037951231002808, "ce_loss": 1.9037951231002808, "logit_z_loss": 99.67067909240723, "muon_lr": 0.06868016007123487, "adamw_lr": 0.003924580575499135, "grad_norm": 0.2757795751094818, "tokens_per_second": 8682488.68444071, "eta_seconds": 3150.442093062401, "data_wait_seconds": 1.8537044525146484e-05}
|
| 1939 |
+
{"step": 19380, "total_steps": 22439, "tokens": 172731924480, "source": "fineweb_hq", "phase": 2, "loss": 1.930611789226532, "ce_loss": 1.930611789226532, "logit_z_loss": 102.77549362182617, "muon_lr": 0.0685897786746317, "adamw_lr": 0.003919415924264669, "grad_norm": 0.37435832619667053, "tokens_per_second": 10160507.672765262, "eta_seconds": 2683.384506177902, "data_wait_seconds": 1.7583370208740234e-05}
|
| 1940 |
+
{"step": 19390, "total_steps": 22439, "tokens": 172821053440, "source": "fineweb_hq", "phase": 2, "loss": 1.9116033911705017, "ce_loss": 1.9116033911705017, "logit_z_loss": 103.38683891296387, "muon_lr": 0.06849646778980163, "adamw_lr": 0.00391408387370295, "grad_norm": 0.3588482737541199, "tokens_per_second": 8607059.317463707, "eta_seconds": 3157.340841007233, "data_wait_seconds": 1.704692840576172e-05}
|
| 1941 |
+
{"step": 19400, "total_steps": 22439, "tokens": 172910182400, "source": "fineweb_hq", "phase": 2, "loss": 1.9180322587490082, "ce_loss": 1.9180322587490082, "logit_z_loss": 102.59267234802246, "muon_lr": 0.06840023555473031, "adamw_lr": 0.003908584888841732, "grad_norm": 0.32407161593437195, "tokens_per_second": 9185864.37740006, "eta_seconds": 2948.692668557167, "data_wait_seconds": 0.15834635496139526}
|
| 1942 |
+
{"step": 19410, "total_steps": 22439, "tokens": 172999311360, "source": "fineweb_hq", "phase": 2, "loss": 1.9428526163101196, "ce_loss": 1.9428526163101196, "logit_z_loss": 96.33477020263672, "muon_lr": 0.06830109036218518, "adamw_lr": 0.003902919449267724, "grad_norm": 0.43458402156829834, "tokens_per_second": 8852409.133024588, "eta_seconds": 3049.696594262123, "data_wait_seconds": 1.71661376953125e-05}
|
| 1943 |
+
{"step": 19420, "total_steps": 22439, "tokens": 173088440320, "source": "fineweb_hq", "phase": 2, "loss": 1.9612203538417816, "ce_loss": 1.9612203538417816, "logit_z_loss": 99.41912269592285, "muon_lr": 0.06819904085898337, "adamw_lr": 0.003897088049084764, "grad_norm": 0.38826051354408264, "tokens_per_second": 8820304.44660908, "eta_seconds": 3050.692091965675, "data_wait_seconds": 0.5098279714584351}
|
| 1944 |
+
{"step": 19430, "total_steps": 22439, "tokens": 173177569280, "source": "fineweb_hq", "phase": 2, "loss": 1.9692934453487396, "ce_loss": 1.9692934453487396, "logit_z_loss": 93.93903732299805, "muon_lr": 0.06809409594523767, "adamw_lr": 0.0038910911968707243, "grad_norm": 0.3337586522102356, "tokens_per_second": 9949211.293054841, "eta_seconds": 2695.5809133052826, "data_wait_seconds": 1.615285873413086e-05}
|
| 1945 |
+
{"step": 19440, "total_steps": 22439, "tokens": 173266698240, "source": "fineweb_hq", "phase": 2, "loss": 1.9974434077739716, "ce_loss": 1.9974434077739716, "logit_z_loss": 95.56488609313965, "muon_lr": 0.06798626477358027, "adamw_lr": 0.0038849294156331577, "grad_norm": 0.3681563138961792, "tokens_per_second": 8405053.16452627, "eta_seconds": 3180.202977991104, "data_wait_seconds": 2.092123031616211e-05}
|
| 1946 |
+
{"step": 19450, "total_steps": 22439, "tokens": 173355827200, "source": "fineweb_hq", "phase": 2, "loss": 1.9871771931648254, "ce_loss": 1.9871771931648254, "logit_z_loss": 100.41986656188965, "muon_lr": 0.06787555674836449, "adamw_lr": 0.003878603242763685, "grad_norm": 0.3114509880542755, "tokens_per_second": 9966454.740391435, "eta_seconds": 2673.031367516518, "data_wait_seconds": 2.5212764739990234e-05}
|
| 1947 |
+
{"step": 19460, "total_steps": 22439, "tokens": 173444956160, "source": "fineweb_hq", "phase": 2, "loss": 1.9999180138111115, "ce_loss": 1.9999180138111115, "logit_z_loss": 94.59343338012695, "muon_lr": 0.06776198152484472, "adamw_lr": 0.0038721132299911263, "grad_norm": 0.3073788285255432, "tokens_per_second": 10175163.865516474, "eta_seconds": 2609.443693971634, "data_wait_seconds": 1.8477439880371094e-05}
|
| 1948 |
+
{"step": 19470, "total_steps": 22439, "tokens": 173534085120, "source": "fineweb_hq", "phase": 2, "loss": 1.9917677640914917, "ce_loss": 1.9917677640914917, "logit_z_loss": 97.6484146118164, "muon_lr": 0.06764554900833418, "adamw_lr": 0.003865459943333382, "grad_norm": 0.32041773200035095, "tokens_per_second": 8071573.74880357, "eta_seconds": 3278.4670062541963, "data_wait_seconds": 1.6570091247558594e-05}
|
| 1949 |
+
{"step": 19480, "total_steps": 22439, "tokens": 173623214080, "source": "fineweb_hq", "phase": 2, "loss": 2.007250279188156, "ce_loss": 2.007250279188156, "logit_z_loss": 98.85545539855957, "muon_lr": 0.0675262693533412, "adamw_lr": 0.003858643963048069, "grad_norm": 0.3199550211429596, "tokens_per_second": 10351366.614703353, "eta_seconds": 2547.8045794010163, "data_wait_seconds": 1.722574234008789e-05}
|
| 1950 |
+
{"step": 19490, "total_steps": 22439, "tokens": 173712343040, "source": "fineweb_hq", "phase": 2, "loss": 2.020092725753784, "ce_loss": 2.020092725753784, "logit_z_loss": 97.80499458312988, "muon_lr": 0.0674041529626835, "adamw_lr": 0.003851665883581914, "grad_norm": 0.3661147654056549, "tokens_per_second": 9236175.258416226, "eta_seconds": 2845.7808095455166, "data_wait_seconds": 1.6689300537109375e-05}
|
| 1951 |
+
{"step": 19500, "total_steps": 22439, "tokens": 173801472000, "source": "fineweb_hq", "phase": 2, "loss": 1.9920507669448853, "ce_loss": 1.9920507669448853, "logit_z_loss": 95.1522045135498, "muon_lr": 0.06727921048658096, "adamw_lr": 0.0038445263135189114, "grad_norm": 0.37789759039878845, "tokens_per_second": 9417165.97574693, "eta_seconds": 2781.6225615501403, "data_wait_seconds": 1.5795230865478516e-05}
|
| 1952 |
+
{"step": 19510, "total_steps": 22439, "tokens": 173890600960, "source": "fineweb_hq", "phase": 2, "loss": 2.0078954100608826, "ce_loss": 2.0078954100608826, "logit_z_loss": 96.09622192382812, "muon_lr": 0.06715145282172677, "adamw_lr": 0.0038372258755272437, "grad_norm": 0.39966803789138794, "tokens_per_second": 8420611.420606073, "eta_seconds": 3100.234778690338, "data_wait_seconds": 1.7583370208740234e-05}
|
| 1953 |
+
{"step": 19520, "total_steps": 22439, "tokens": 173979729920, "source": "fineweb_hq", "phase": 2, "loss": 2.0032006204128265, "ce_loss": 2.0032006204128265, "logit_z_loss": 93.53397369384766, "muon_lr": 0.06702089111033711, "adamw_lr": 0.0038297652063049776, "grad_norm": 0.3665102422237396, "tokens_per_second": 10205999.502737228, "eta_seconds": 2549.161737370491, "data_wait_seconds": 1.5914440155029297e-05}
|
| 1954 |
+
{"step": 19530, "total_steps": 22439, "tokens": 174068858880, "source": "fineweb_hq", "phase": 2, "loss": 2.0087883174419403, "ce_loss": 2.0087883174419403, "logit_z_loss": 98.91171455383301, "muon_lr": 0.06688753673917937, "adamw_lr": 0.003822144956524535, "grad_norm": 0.3983410596847534, "tokens_per_second": 7890800.593682166, "eta_seconds": 3285.8027719974516, "data_wait_seconds": 1.9490718841552734e-05}
|
| 1955 |
+
{"step": 19540, "total_steps": 22439, "tokens": 174157987840, "source": "fineweb_hq", "phase": 2, "loss": 2.004176527261734, "ce_loss": 2.004176527261734, "logit_z_loss": 95.4532470703125, "muon_lr": 0.06675140133857907, "adamw_lr": 0.0038143657907759466, "grad_norm": 0.3351048529148102, "tokens_per_second": 10030181.004547568, "eta_seconds": 2576.0737011909487, "data_wait_seconds": 1.7821788787841797e-05}
|
| 1956 |
+
{"step": 19550, "total_steps": 22439, "tokens": 174247116800, "source": "fineweb_hq", "phase": 2, "loss": 1.9885312616825104, "ce_loss": 1.9885312616825104, "logit_z_loss": 92.02595138549805, "muon_lr": 0.06661249678140556, "adamw_lr": 0.0038064283875088884, "grad_norm": 0.3306865096092224, "tokens_per_second": 8573899.901903128, "eta_seconds": 3003.2257010936733, "data_wait_seconds": 1.4781951904296875e-05}
|
| 1957 |
+
{"step": 19560, "total_steps": 22439, "tokens": 174336245760, "source": "fineweb_hq", "phase": 2, "loss": 1.9839977025985718, "ce_loss": 1.9839977025985718, "logit_z_loss": 101.25559997558594, "muon_lr": 0.06647083518203649, "adamw_lr": 0.0037983334389735134, "grad_norm": 0.34344813227653503, "tokens_per_second": 9498528.902057337, "eta_seconds": 2701.4949208021167, "data_wait_seconds": 2.0742416381835938e-05}
|
| 1958 |
+
{"step": 19570, "total_steps": 22439, "tokens": 174425374720, "source": "fineweb_hq", "phase": 2, "loss": 2.0125721991062164, "ce_loss": 2.0125721991062164, "logit_z_loss": 97.73648452758789, "muon_lr": 0.06632642889530135, "adamw_lr": 0.0037900816511600775, "grad_norm": 0.32962390780448914, "tokens_per_second": 8376401.411692941, "eta_seconds": 3052.754681539536, "data_wait_seconds": 1.8894672393798828e-05}
|
| 1959 |
+
{"step": 19580, "total_steps": 22439, "tokens": 174514503680, "source": "fineweb_hq", "phase": 2, "loss": 2.00698584318161, "ce_loss": 2.00698584318161, "logit_z_loss": 93.60843276977539, "muon_lr": 0.06617929051540392, "adamw_lr": 0.0037816737437373666, "grad_norm": 0.2984445095062256, "tokens_per_second": 8812773.009522965, "eta_seconds": 2891.481448173523, "data_wait_seconds": 1.8298625946044922e-05}
|
| 1960 |
+
{"step": 19590, "total_steps": 22439, "tokens": 174603632640, "source": "fineweb_hq", "phase": 2, "loss": 2.0014726519584656, "ce_loss": 2.0014726519584656, "logit_z_loss": 97.14739799499512, "muon_lr": 0.06602943287482382, "adamw_lr": 0.0037731104499899326, "grad_norm": 0.312802255153656, "tokens_per_second": 10996148.945512442, "eta_seconds": 2309.248522353172, "data_wait_seconds": 1.9311904907226562e-05}
|
| 1961 |
+
{"step": 19600, "total_steps": 22439, "tokens": 174692761600, "source": "fineweb_hq", "phase": 2, "loss": 1.9961905479431152, "ce_loss": 1.9961905479431152, "logit_z_loss": 95.27328300476074, "muon_lr": 0.06587686904319746, "adamw_lr": 0.003764392516754141, "grad_norm": 0.32714709639549255, "tokens_per_second": 8451840.987512216, "eta_seconds": 2993.8698304176332, "data_wait_seconds": 1.537799835205078e-05}
|
| 1962 |
+
{"step": 19610, "total_steps": 22439, "tokens": 174781890560, "source": "fineweb_hq", "phase": 2, "loss": 2.0066860914230347, "ce_loss": 2.0066860914230347, "logit_z_loss": 95.27567481994629, "muon_lr": 0.06572161232617814, "adamw_lr": 0.0037555207043530364, "grad_norm": 0.3989483118057251, "tokens_per_second": 9394212.591067966, "eta_seconds": 2684.05494761467, "data_wait_seconds": 1.7523765563964844e-05}
|
| 1963 |
+
{"step": 19620, "total_steps": 22439, "tokens": 174871019520, "source": "fineweb_hq", "phase": 2, "loss": 1.991454839706421, "ce_loss": 1.991454839706421, "logit_z_loss": 91.48034286499023, "muon_lr": 0.06556367626427555, "adamw_lr": 0.003746495786530031, "grad_norm": 0.2970179617404938, "tokens_per_second": 8303240.967105117, "eta_seconds": 3025.981532216072, "data_wait_seconds": 2.0205974578857422e-05}
|
| 1964 |
+
{"step": 19630, "total_steps": 22439, "tokens": 174960148480, "source": "fineweb_hq", "phase": 2, "loss": 1.981344848871231, "ce_loss": 1.981344848871231, "logit_z_loss": 98.25697326660156, "muon_lr": 0.06540307463167494, "adamw_lr": 0.0037373185503814252, "grad_norm": 0.35564327239990234, "tokens_per_second": 10067382.412217127, "eta_seconds": 2486.875320601463, "data_wait_seconds": 2.1219253540039062e-05}
|
| 1965 |
+
{"step": 19640, "total_steps": 22439, "tokens": 175049277440, "source": "fineweb_hq", "phase": 2, "loss": 1.9723089039325714, "ce_loss": 1.9723089039325714, "logit_z_loss": 98.7834415435791, "muon_lr": 0.0652398214350358, "adamw_lr": 0.00372798979628776, "grad_norm": 0.33393755555152893, "tokens_per_second": 9085774.006620882, "eta_seconds": 2745.7425075531005, "data_wait_seconds": 2.092123031616211e-05}
|
| 1966 |
+
{"step": 19650, "total_steps": 22439, "tokens": 175138406400, "source": "fineweb_hq", "phase": 2, "loss": 2.0165835916996, "ce_loss": 2.0165835916996, "logit_z_loss": 93.70169067382812, "muon_lr": 0.06507393091227028, "adamw_lr": 0.0037185103378440153, "grad_norm": 0.4647543132305145, "tokens_per_second": 9257890.543828143, "eta_seconds": 2685.0681401252746, "data_wait_seconds": 2.181529998779297e-05}
|
| 1967 |
+
{"step": 19660, "total_steps": 22439, "tokens": 175227535360, "source": "fineweb_hq", "phase": 2, "loss": 1.9912970960140228, "ce_loss": 1.9912970960140228, "logit_z_loss": 90.95748710632324, "muon_lr": 0.06490541753130141, "adamw_lr": 0.003708881001788652, "grad_norm": 0.3512067496776581, "tokens_per_second": 9456474.21767216, "eta_seconds": 2619.2571791410446, "data_wait_seconds": 1.7821788787841797e-05}
|
| 1968 |
+
{"step": 19670, "total_steps": 22439, "tokens": 175316664320, "source": "fineweb_hq", "phase": 2, "loss": 1.971030205488205, "ce_loss": 1.971030205488205, "logit_z_loss": 98.85235595703125, "muon_lr": 0.06473429598880139, "adamw_lr": 0.0036991026279315077, "grad_norm": 0.3620174825191498, "tokens_per_second": 9874667.653546447, "eta_seconds": 2499.305281949043, "data_wait_seconds": 2.396106719970703e-05}
|
| 1969 |
+
{"step": 19680, "total_steps": 22439, "tokens": 175405793280, "source": "fineweb_hq", "phase": 2, "loss": 1.98936727643013, "ce_loss": 1.98936727643013, "logit_z_loss": 96.70642852783203, "muon_lr": 0.06456058120890974, "adamw_lr": 0.003689176069080556, "grad_norm": 0.29864001274108887, "tokens_per_second": 8887392.261896458, "eta_seconds": 2766.9173745632174, "data_wait_seconds": 1.823902130126953e-05}
|
| 1970 |
+
{"step": 19690, "total_steps": 22439, "tokens": 175494922240, "source": "fineweb_hq", "phase": 2, "loss": 1.9724730551242828, "ce_loss": 1.9724730551242828, "logit_z_loss": 100.79594230651855, "muon_lr": 0.0643842883419318, "adamw_lr": 0.0036791021909675314, "grad_norm": 0.3698233664035797, "tokens_per_second": 10028661.539239341, "eta_seconds": 2443.1526588201523, "data_wait_seconds": 1.7464160919189453e-05}
|
| 1971 |
+
{"step": 19700, "total_steps": 22439, "tokens": 175584051200, "source": "fineweb_hq", "phase": 2, "loss": 2.0018686056137085, "ce_loss": 2.0018686056137085, "logit_z_loss": 97.60370254516602, "muon_lr": 0.06420543276301735, "adamw_lr": 0.00366888187217242, "grad_norm": 0.27720940113067627, "tokens_per_second": 8800866.81945009, "eta_seconds": 2773.86564810276, "data_wait_seconds": 0.16173052787780762}
|
| 1972 |
+
{"step": 19710, "total_steps": 22439, "tokens": 175673180160, "source": "fineweb_hq", "phase": 2, "loss": 1.9875790774822235, "ce_loss": 1.9875790774822235, "logit_z_loss": 96.14850807189941, "muon_lr": 0.0640240300708197, "adamw_lr": 0.0036585160040468402, "grad_norm": 0.35521429777145386, "tokens_per_second": 9749814.102533013, "eta_seconds": 2494.7443026304245, "data_wait_seconds": 1.823902130126953e-05}
|
| 1973 |
+
{"step": 19720, "total_steps": 22439, "tokens": 175762309120, "source": "fineweb_hq", "phase": 2, "loss": 1.978615790605545, "ce_loss": 1.978615790605545, "logit_z_loss": 96.73612785339355, "muon_lr": 0.06384009608613532, "adamw_lr": 0.003648005490636304, "grad_norm": 0.30277007818222046, "tokens_per_second": 9652237.643871957, "eta_seconds": 2510.730166220665, "data_wait_seconds": 0.040255069732666016}
|
| 1974 |
+
{"step": 19730, "total_steps": 22439, "tokens": 175851438080, "source": "fineweb_hq", "phase": 2, "loss": 1.9813168942928314, "ce_loss": 1.9813168942928314, "logit_z_loss": 100.83143043518066, "muon_lr": 0.06365364685052398, "adamw_lr": 0.00363735124860137, "grad_norm": 0.3035475015640259, "tokens_per_second": 8569255.159907652, "eta_seconds": 2817.6352335691454, "data_wait_seconds": 1.8358230590820312e-05}
|
| 1975 |
+
{"step": 19740, "total_steps": 22439, "tokens": 175940567040, "source": "fineweb_hq", "phase": 2, "loss": 1.9735566973686218, "ce_loss": 1.9735566973686218, "logit_z_loss": 97.95806694030762, "muon_lr": 0.06346469862490975, "adamw_lr": 0.0036265542071376995, "grad_norm": 0.3591097295284271, "tokens_per_second": 10294995.991535626, "eta_seconds": 2336.6600942611694, "data_wait_seconds": 1.5735626220703125e-05}
|
| 1976 |
+
{"step": 19750, "total_steps": 22439, "tokens": 176029696000, "source": "fineweb_hq", "phase": 2, "loss": 1.979945957660675, "ce_loss": 1.979945957660675, "logit_z_loss": 99.81876373291016, "muon_lr": 0.0632732678881628, "adamw_lr": 0.003615615307895017, "grad_norm": 0.266200453042984, "tokens_per_second": 8797283.656927874, "eta_seconds": 2724.338361549377, "data_wait_seconds": 1.6987323760986328e-05}
|
| 1977 |
+
{"step": 19760, "total_steps": 22439, "tokens": 176118824960, "source": "fineweb_hq", "phase": 2, "loss": 1.9740092158317566, "ce_loss": 1.9740092158317566, "logit_z_loss": 93.13092422485352, "muon_lr": 0.06307937133566228, "adamw_lr": 0.0036045355048949877, "grad_norm": 0.3467254340648651, "tokens_per_second": 9031150.276670178, "eta_seconds": 2643.921056842804, "data_wait_seconds": 1.7464160919189453e-05}
|
| 1978 |
+
{"step": 19770, "total_steps": 22439, "tokens": 176207953920, "source": "fineweb_hq", "phase": 2, "loss": 1.9817664623260498, "ce_loss": 1.9817664623260498, "logit_z_loss": 96.25656127929688, "muon_lr": 0.06288302587784018, "adamw_lr": 0.0035933157644480102, "grad_norm": 0.341808944940567, "tokens_per_second": 8288598.884720207, "eta_seconds": 2870.0290308237077, "data_wait_seconds": 0.3086075782775879}
|
| 1979 |
+
{"step": 19780, "total_steps": 22439, "tokens": 176297082880, "source": "fineweb_hq", "phase": 2, "loss": 1.9727942049503326, "ce_loss": 1.9727942049503326, "logit_z_loss": 100.89604568481445, "muon_lr": 0.06268424863870652, "adamw_lr": 0.003581957065068943, "grad_norm": 0.2886621057987213, "tokens_per_second": 8984774.772270946, "eta_seconds": 2637.7278301000597, "data_wait_seconds": 0.23426759243011475}
|
| 1980 |
+
{"step": 19790, "total_steps": 22439, "tokens": 176386211840, "source": "fineweb_hq", "phase": 2, "loss": 1.9756250381469727, "ce_loss": 1.9756250381469727, "logit_z_loss": 99.3545036315918, "muon_lr": 0.06248305695435593, "adamw_lr": 0.003570460397391767, "grad_norm": 0.3333895206451416, "tokens_per_second": 7230756.135784593, "eta_seconds": 3265.2548448085786, "data_wait_seconds": 0.6596333384513855}
|
| 1981 |
+
{"step": 19800, "total_steps": 22439, "tokens": 176475340800, "source": "fineweb_hq", "phase": 2, "loss": 1.9855551719665527, "ce_loss": 1.9855551719665527, "logit_z_loss": 101.0711669921875, "muon_lr": 0.06227946837145571, "adamw_lr": 0.003558826764083183, "grad_norm": 0.3963225781917572, "tokens_per_second": 8723614.971150862, "eta_seconds": 2696.259821391106, "data_wait_seconds": 1.9609928131103516e-05}
|
| 1982 |
+
{"step": 19810, "total_steps": 22439, "tokens": 176564469760, "source": "fineweb_hq", "phase": 2, "loss": 1.9546213150024414, "ce_loss": 1.9546213150024414, "logit_z_loss": 99.74259757995605, "muon_lr": 0.06207350064571549, "adamw_lr": 0.0035470571797551705, "grad_norm": 0.3589343726634979, "tokens_per_second": 8491655.581016244, "eta_seconds": 2759.4152118444445, "data_wait_seconds": 1.919269561767578e-05}
|
| 1983 |
+
{"step": 19820, "total_steps": 22439, "tokens": 176653598720, "source": "fineweb_hq", "phase": 2, "loss": 1.9730973839759827, "ce_loss": 1.9730973839759827, "logit_z_loss": 99.34999465942383, "muon_lr": 0.0618651717403387, "adamw_lr": 0.003535152670876497, "grad_norm": 0.3444134593009949, "tokens_per_second": 9766353.595522787, "eta_seconds": 2390.132038092613, "data_wait_seconds": 3.0934810638427734e-05}
|
| 1984 |
+
{"step": 19830, "total_steps": 22439, "tokens": 176742727680, "source": "fineweb_hq", "phase": 2, "loss": 1.952784240245819, "ce_loss": 1.952784240245819, "logit_z_loss": 99.39868354797363, "muon_lr": 0.06165449982445595, "adamw_lr": 0.0035231142756831967, "grad_norm": 0.2932511270046234, "tokens_per_second": 9002558.980823714, "eta_seconds": 2583.0150864362718, "data_wait_seconds": 1.7523765563964844e-05}
|
| 1985 |
+
{"step": 19840, "total_steps": 22439, "tokens": 176831856640, "source": "fineweb_hq", "phase": 2, "loss": 1.9490482807159424, "ce_loss": 1.9490482807159424, "logit_z_loss": 96.5870475769043, "muon_lr": 0.061441503271540396, "adamw_lr": 0.003510943044088022, "grad_norm": 0.2961852252483368, "tokens_per_second": 7902561.210777192, "eta_seconds": 2931.2796302556994, "data_wait_seconds": 2.1755695343017578e-05}
|
| 1986 |
+
{"step": 19850, "total_steps": 22439, "tokens": 176920985600, "source": "fineweb_hq", "phase": 2, "loss": 1.9600580036640167, "ce_loss": 1.9600580036640167, "logit_z_loss": 97.35949516296387, "muon_lr": 0.061226200657805374, "adamw_lr": 0.003498640037588878, "grad_norm": 0.3256329298019409, "tokens_per_second": 8332249.810788183, "eta_seconds": 2769.4186165809633, "data_wait_seconds": 2.288818359375e-05}
|
| 1987 |
+
{"step": 19860, "total_steps": 22439, "tokens": 177010114560, "source": "fineweb_hq", "phase": 2, "loss": 1.947264403104782, "ce_loss": 1.947264403104782, "logit_z_loss": 97.30531311035156, "muon_lr": 0.06100861076058425, "adamw_lr": 0.003486206329176243, "grad_norm": 0.31083858013153076, "tokens_per_second": 8613086.2194345, "eta_seconds": 2668.7714715003967, "data_wait_seconds": 0.160763680934906}
|
| 1988 |
+
{"step": 19870, "total_steps": 22439, "tokens": 177099243520, "source": "fineweb_hq", "phase": 2, "loss": 1.961327999830246, "ce_loss": 1.961327999830246, "logit_z_loss": 98.70391464233398, "muon_lr": 0.06078875255669283, "adamw_lr": 0.0034736430032395902, "grad_norm": 0.34762781858444214, "tokens_per_second": 8662530.860217085, "eta_seconds": 2643.24943754673, "data_wait_seconds": 1.7404556274414062e-05}
|
| 1989 |
+
{"step": 19880, "total_steps": 22439, "tokens": 177188372480, "source": "fineweb_hq", "phase": 2, "loss": 1.9673119485378265, "ce_loss": 1.9673119485378265, "logit_z_loss": 98.9455680847168, "muon_lr": 0.06056664522077426, "adamw_lr": 0.0034609511554728146, "grad_norm": 0.3625636100769043, "tokens_per_second": 8940509.385865271, "eta_seconds": 2551.0963502883915, "data_wait_seconds": 1.895427703857422e-05}
|
| 1990 |
+
{"step": 19890, "total_steps": 22439, "tokens": 177277501440, "source": "fineweb_hq", "phase": 2, "loss": 1.9690393507480621, "ce_loss": 1.9690393507480621, "logit_z_loss": 94.27217483520508, "muon_lr": 0.06034230812362677, "adamw_lr": 0.0034481318927786723, "grad_norm": 0.29470884799957275, "tokens_per_second": 9196279.624652686, "eta_seconds": 2470.4524907112122, "data_wait_seconds": 1.996755599975586e-05}
|
| 1991 |
+
{"step": 19900, "total_steps": 22439, "tokens": 177366630400, "source": "fineweb_hq", "phase": 2, "loss": 1.9784173965454102, "ce_loss": 1.9784173965454102, "logit_z_loss": 99.94176292419434, "muon_lr": 0.060115760830514314, "adamw_lr": 0.003435186333172246, "grad_norm": 0.2931779623031616, "tokens_per_second": 9354130.047426987, "eta_seconds": 2419.235442447662, "data_wait_seconds": 1.806020736694336e-05}
|
| 1992 |
+
{"step": 19910, "total_steps": 22439, "tokens": 177455759360, "source": "fineweb_hq", "phase": 2, "loss": 1.979276180267334, "ce_loss": 1.979276180267334, "logit_z_loss": 95.98761367797852, "muon_lr": 0.05988702309946011, "adamw_lr": 0.0034221156056834347, "grad_norm": 0.324918657541275, "tokens_per_second": 9239566.563889902, "eta_seconds": 2439.585648107529, "data_wait_seconds": 1.9431114196777344e-05}
|
| 1993 |
+
{"step": 19920, "total_steps": 22439, "tokens": 177544888320, "source": "fineweb_hq", "phase": 2, "loss": 1.9856651723384857, "ce_loss": 1.9856651723384857, "logit_z_loss": 100.09841728210449, "muon_lr": 0.05965611487952359, "adamw_lr": 0.0034089208502584905, "grad_norm": 0.3025656044483185, "tokens_per_second": 8879594.851809626, "eta_seconds": 2528.4470067262646, "data_wait_seconds": 0.16344988346099854}
|
| 1994 |
+
{"step": 19930, "total_steps": 22439, "tokens": 177634017280, "source": "fineweb_hq", "phase": 2, "loss": 1.9656601250171661, "ce_loss": 1.9656601250171661, "logit_z_loss": 99.07984924316406, "muon_lr": 0.05942305630906046, "adamw_lr": 0.0033956032176605975, "grad_norm": 0.32972630858421326, "tokens_per_second": 9167879.43925435, "eta_seconds": 2439.2179469823836, "data_wait_seconds": 1.633167266845703e-05}
|
| 1995 |
+
{"step": 19940, "total_steps": 22439, "tokens": 177723146240, "source": "fineweb_hq", "phase": 2, "loss": 1.9596152901649475, "ce_loss": 1.9596152901649475, "logit_z_loss": 99.73912239074707, "muon_lr": 0.05918786771396645, "adamw_lr": 0.003382163869369511, "grad_norm": 0.2992822229862213, "tokens_per_second": 8360275.967007774, "eta_seconds": 2664.1856311798097, "data_wait_seconds": 1.901388168334961e-05}
|
| 1996 |
+
{"step": 19950, "total_steps": 22439, "tokens": 177812275200, "source": "fineweb_hq", "phase": 2, "loss": 1.9457648694515228, "ce_loss": 1.9457648694515228, "logit_z_loss": 100.90079116821289, "muon_lr": 0.05895056960590455, "adamw_lr": 0.00336860397748026, "grad_norm": 0.3093985915184021, "tokens_per_second": 7604292.01620515, "eta_seconds": 2917.3259123563766, "data_wait_seconds": 1.633167266845703e-05}
|
| 1997 |
+
{"step": 19960, "total_steps": 22439, "tokens": 177901404160, "source": "fineweb_hq", "phase": 2, "loss": 1.9273811876773834, "ce_loss": 1.9273811876773834, "logit_z_loss": 100.26827049255371, "muon_lr": 0.058711182680516165, "adamw_lr": 0.0033549247246009235, "grad_norm": 0.316666841506958, "tokens_per_second": 6585183.510419693, "eta_seconds": 3355.2700769901276, "data_wait_seconds": 2.0682811737060547e-05}
|
| 1998 |
+
{"step": 19970, "total_steps": 22439, "tokens": 177990533120, "source": "fineweb_hq", "phase": 2, "loss": 1.9429328739643097, "ce_loss": 1.9429328739643097, "logit_z_loss": 96.3702449798584, "muon_lr": 0.05846972781561611, "adamw_lr": 0.0033411273037494914, "grad_norm": 0.2850782573223114, "tokens_per_second": 6293007.523700372, "eta_seconds": 3496.8876393556598, "data_wait_seconds": 2.0384788513183594e-05}
|
| 1999 |
+
{"step": 19980, "total_steps": 22439, "tokens": 178079662080, "source": "fineweb_hq", "phase": 2, "loss": 1.9387704730033875, "ce_loss": 1.9387704730033875, "logit_z_loss": 100.60917472839355, "muon_lr": 0.058226226069371854, "adamw_lr": 0.00332721291824982, "grad_norm": 0.34937146306037903, "tokens_per_second": 7941680.842040825, "eta_seconds": 2759.7194724798205, "data_wait_seconds": 1.8835067749023438e-05}
|
| 2000 |
+
{"step": 19990, "total_steps": 22439, "tokens": 178168791040, "source": "fineweb_hq", "phase": 2, "loss": 1.9374864399433136, "ce_loss": 1.9374864399433136, "logit_z_loss": 101.71585464477539, "muon_lr": 0.057980698678466905, "adamw_lr": 0.00331318278162668, "grad_norm": 0.34407687187194824, "tokens_per_second": 5834007.040807811, "eta_seconds": 3741.4562840461726, "data_wait_seconds": 0.41281574964523315}
|
| 2001 |
+
{"step": 20000, "total_steps": 22439, "tokens": 178257920000, "source": "fineweb_hq", "phase": 2, "loss": 1.9428208768367767, "ce_loss": 1.9428208768367767, "logit_z_loss": 100.47715759277344, "muon_lr": 0.05773316705624869, "adamw_lr": 0.003299038117499925, "grad_norm": 0.365927129983902, "tokens_per_second": 7364924.441888035, "eta_seconds": 2951.6329075098038, "data_wait_seconds": 2.014636993408203e-05}
|
| 2002 |
+
{"step": 20010, "total_steps": 22439, "tokens": 178347048960, "source": "fineweb_hq", "phase": 2, "loss": 1.9283939599990845, "ce_loss": 1.9283939599990845, "logit_z_loss": 100.78992462158203, "muon_lr": 0.057483652790861046, "adamw_lr": 0.0032847801594777736, "grad_norm": 0.3201388418674469, "tokens_per_second": 11306903.475998694, "eta_seconds": 1914.7085167884827, "data_wait_seconds": 1.8596649169921875e-05}
|
| 2003 |
+
{"step": 20020, "total_steps": 22439, "tokens": 178436177920, "source": "fineweb_hq", "phase": 2, "loss": 1.9457659423351288, "ce_loss": 1.9457659423351288, "logit_z_loss": 102.80153274536133, "muon_lr": 0.05723217764336141, "adamw_lr": 0.003270410151049223, "grad_norm": 0.33229416608810425, "tokens_per_second": 7529758.596148631, "eta_seconds": 2863.344840168953, "data_wait_seconds": 0.2085026502609253}
|
| 2004 |
+
{"step": 20030, "total_steps": 22439, "tokens": 178525306880, "source": "fineweb_hq", "phase": 2, "loss": 1.9406559467315674, "ce_loss": 1.9406559467315674, "logit_z_loss": 99.24288749694824, "muon_lr": 0.05697876354582294, "adamw_lr": 0.0032559293454755963, "grad_norm": 0.31985676288604736, "tokens_per_second": 7981005.170398412, "eta_seconds": 2690.283492565155, "data_wait_seconds": 1.8715858459472656e-05}
|
| 2005 |
+
{"step": 20040, "total_steps": 22439, "tokens": 178614435840, "source": "fineweb_hq", "phase": 2, "loss": 1.9395206272602081, "ce_loss": 1.9395206272602081, "logit_z_loss": 103.45383262634277, "muon_lr": 0.0567234325994218, "adamw_lr": 0.0032413390056812456, "grad_norm": 0.29466670751571655, "tokens_per_second": 8704839.06775217, "eta_seconds": 2456.339208292961, "data_wait_seconds": 2.7477741241455078e-05}
|
| 2006 |
+
{"step": 20050, "total_steps": 22439, "tokens": 178703564800, "source": "fineweb_hq", "phase": 2, "loss": 1.9036086797714233, "ce_loss": 1.9036086797714233, "logit_z_loss": 99.11339569091797, "muon_lr": 0.05646620707250958, "adamw_lr": 0.0032266404041434043, "grad_norm": 0.3292168974876404, "tokens_per_second": 6681604.841558023, "eta_seconds": 3186.795545220375, "data_wait_seconds": 1.8835067749023438e-05}
|
| 2007 |
+
{"step": 20060, "total_steps": 22439, "tokens": 178792693760, "source": "fineweb_hq", "phase": 2, "loss": 1.9195380508899689, "ce_loss": 1.9195380508899689, "logit_z_loss": 98.8975944519043, "muon_lr": 0.05620710939867121, "adamw_lr": 0.003211834822781212, "grad_norm": 0.36700335144996643, "tokens_per_second": 9213887.538765796, "eta_seconds": 2301.2848262786865, "data_wait_seconds": 2.0325183868408203e-05}
|
| 2008 |
+
{"step": 20070, "total_steps": 22439, "tokens": 178881822720, "source": "fineweb_hq", "phase": 2, "loss": 1.8926385939121246, "ce_loss": 1.8926385939121246, "logit_z_loss": 100.59559631347656, "muon_lr": 0.05594616217476845, "adamw_lr": 0.0031969235528439112, "grad_norm": 0.36090517044067383, "tokens_per_second": 7921955.218737889, "eta_seconds": 2665.3332467794417, "data_wait_seconds": 1.8537044525146484e-05}
|
| 2009 |
+
{"step": 20080, "total_steps": 22439, "tokens": 178970951680, "source": "fineweb_hq", "phase": 2, "loss": 1.8612332940101624, "ce_loss": 1.8612332940101624, "logit_z_loss": 106.54780006408691, "muon_lr": 0.05568338815896912, "adamw_lr": 0.0031819078947982354, "grad_norm": 0.30154821276664734, "tokens_per_second": 8663289.282040535, "eta_seconds": 2426.967515397072, "data_wait_seconds": 1.8715858459472656e-05}
|
| 2010 |
+
{"step": 20090, "total_steps": 22439, "tokens": 179060080640, "source": "fineweb_hq", "phase": 2, "loss": 1.856196403503418, "ce_loss": 1.856196403503418, "logit_z_loss": 102.11246299743652, "muon_lr": 0.0554188102687623, "adamw_lr": 0.0031667891582149886, "grad_norm": 0.28575363755226135, "tokens_per_second": 9757163.273537377, "eta_seconds": 2145.7458604574203, "data_wait_seconds": 1.811981201171875e-05}
|
| 2011 |
+
{"step": 20100, "total_steps": 22439, "tokens": 179149209600, "source": "fineweb_hq", "phase": 2, "loss": 1.8791177868843079, "ce_loss": 1.8791177868843079, "logit_z_loss": 104.77649116516113, "muon_lr": 0.05515245157895957, "adamw_lr": 0.0031515686616548325, "grad_norm": 0.30561697483062744, "tokens_per_second": 8732173.920039145, "eta_seconds": 2387.408214139938, "data_wait_seconds": 1.9311904907226562e-05}
|
| 2012 |
+
{"step": 20110, "total_steps": 22439, "tokens": 179238338560, "source": "fineweb_hq", "phase": 2, "loss": 1.8546561896800995, "ce_loss": 1.8546561896800995, "logit_z_loss": 105.10050582885742, "muon_lr": 0.05488433531968263, "adamw_lr": 0.0031362477325532928, "grad_norm": 0.3312560021877289, "tokens_per_second": 9245451.324482035, "eta_seconds": 2245.226766705513, "data_wait_seconds": 1.9252300262451172e-05}
|
| 2013 |
+
{"step": 20120, "total_steps": 22439, "tokens": 179327467520, "source": "fineweb_hq", "phase": 2, "loss": 1.876363068819046, "ce_loss": 1.876363068819046, "logit_z_loss": 103.58477401733398, "muon_lr": 0.05461448487433723, "adamw_lr": 0.0031208277071049844, "grad_norm": 0.29391342401504517, "tokens_per_second": 9727232.970499212, "eta_seconds": 2124.8597506284714, "data_wait_seconds": 1.8894672393798828e-05}
|
| 2014 |
+
{"step": 20130, "total_steps": 22439, "tokens": 179416596480, "source": "fineweb_hq", "phase": 2, "loss": 1.8685232996940613, "ce_loss": 1.8685232996940613, "logit_z_loss": 103.49335098266602, "muon_lr": 0.05434292377757396, "adamw_lr": 0.003105309930147083, "grad_norm": 0.30132758617401123, "tokens_per_second": 8185050.319996767, "eta_seconds": 2514.3250266551972, "data_wait_seconds": 1.7762184143066406e-05}
|
| 2015 |
+
{"step": 20140, "total_steps": 22439, "tokens": 179505725440, "source": "fineweb_hq", "phase": 2, "loss": 1.8603698015213013, "ce_loss": 1.8603698015213013, "logit_z_loss": 105.01471710205078, "muon_lr": 0.054069675713235524, "adamw_lr": 0.0030896957550420297, "grad_norm": 0.29002025723457336, "tokens_per_second": 9732955.231270514, "eta_seconds": 2105.2956082820892, "data_wait_seconds": 1.817941665649414e-05}
|
| 2016 |
+
{"step": 20150, "total_steps": 22439, "tokens": 179594854400, "source": "fineweb_hq", "phase": 2, "loss": 1.8466525077819824, "ce_loss": 1.8466525077819824, "logit_z_loss": 103.0981616973877, "muon_lr": 0.053794764512291336, "adamw_lr": 0.0030739865435595046, "grad_norm": 0.2753201127052307, "tokens_per_second": 5952425.325244545, "eta_seconds": 3427.446432209015, "data_wait_seconds": 0.5249242186546326}
|
| 2017 |
+
{"step": 20160, "total_steps": 22439, "tokens": 179683983360, "source": "fineweb_hq", "phase": 2, "loss": 1.8820434212684631, "ce_loss": 1.8820434212684631, "logit_z_loss": 107.19742965698242, "muon_lr": 0.053518214150759046, "adamw_lr": 0.0030581836657576594, "grad_norm": 0.2873774766921997, "tokens_per_second": 9461497.907709168, "eta_seconds": 2146.8577367067337, "data_wait_seconds": 1.6033649444580078e-05}
|
| 2018 |
+
{"step": 20170, "total_steps": 22439, "tokens": 179773112320, "source": "fineweb_hq", "phase": 2, "loss": 1.8838628828525543, "ce_loss": 1.8838628828525543, "logit_z_loss": 100.98152160644531, "muon_lr": 0.05324004874761351, "adamw_lr": 0.003042288499863629, "grad_norm": 0.28046002984046936, "tokens_per_second": 8796993.627367357, "eta_seconds": 2298.894586110115, "data_wait_seconds": 1.9073486328125e-05}
|
| 2019 |
+
{"step": 20180, "total_steps": 22439, "tokens": 179862241280, "source": "fineweb_hq", "phase": 2, "loss": 1.9100408852100372, "ce_loss": 1.9100408852100372, "logit_z_loss": 104.36373710632324, "muon_lr": 0.05296029256268338, "adamw_lr": 0.0030263024321533355, "grad_norm": 0.2785077691078186, "tokens_per_second": 8765321.587364828, "eta_seconds": 2297.0328998565674, "data_wait_seconds": 1.704692840576172e-05}
|
| 2020 |
+
{"step": 20190, "total_steps": 22439, "tokens": 179951370240, "source": "fineweb_hq", "phase": 2, "loss": 1.9417555928230286, "ce_loss": 1.9417555928230286, "logit_z_loss": 102.40350532531738, "muon_lr": 0.05267896999453518, "adamw_lr": 0.0030102268568305817, "grad_norm": 0.30016791820526123, "tokens_per_second": 8925125.557419682, "eta_seconds": 2245.91833190918, "data_wait_seconds": 1.704692840576172e-05}
|