Hugging Face's logo

mpashkovskii
/
dspark-qwen3-8b

Safetensors
speculators
dspark
speculative-decoding
draft-model
custom_code
Model card Files
xet
 Community
mpashkovskii commited on
Commit
bc9dbf5
·
verified ·
1 Parent(s): 8aef8aa

v7: DSpark speculator for Qwen/Qwen3-8B

Browse files
Files changed (5) hide show
  1. model.safetensors +1 -1
  2. optimizer_state_dict.pt +1 -1
  3. scheduler_state_dict.pt +1 -1
  4. training_state.json +1 -1
  5. val_metrics.json +1 -1
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b20e89945730a7ca568a8ca2c1972cfd080a8b65167cfe8d5e78e594d877dec6
3
  size 3698655482
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e074dba3c29f1ea0456c33818b6e243f46a46909aaef7437f65136daf0b253fc
3
  size 3698655482
optimizer_state_dict.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:818fe93e6c1cdff9e4309078807135242c72f23d34959cf785ce6d472936ddfc
3
  size 4382932183
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e822a3c604a8e1219246ea53ade9b0163af50fd674cc0eb6e2650c94d612c42
3
  size 4382932183
scheduler_state_dict.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:391a8af730e464ea274b496f17a92a956edec037ab0a897a8548d5f88dc80d66
3
  size 1531
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:18ee8bc4a8cc6564752d77740f97fa7c2fdcb0b2eaaf33478ead4cad01a2b4f6
3
  size 1531
training_state.json CHANGED
@@ -1 +1 @@
1
- {"epoch": 2, "local_step": 0, "global_step": 10733}
 
1
+ {"epoch": 4, "local_step": 0, "global_step": 17895}
val_metrics.json CHANGED
@@ -1 +1 @@
1
- {"confidence_loss_epoch": 0.3996266642677222, "confidence_abs_error_epoch": 0.02379905994113253, "confidence_pred_mean_epoch": 0.09513526574331277, "confidence_cumprod_bias_epoch": 0.00793046181361493, "loss_epoch": 0.016264476283496487, "ce_loss_epoch": 0.026335223726050332, "tv_loss_epoch": 0.006086843472545549, "accept_rate_epoch": 0.6540066899581706, "accept_len_epoch": 2.2047277528849136, "full_acc_epoch": 0.6681219985246512, "position_1_acc_epoch": 0.744284980053469, "position_2_acc_epoch": 0.5918931619895018}
 
1
+ {"confidence_loss_epoch": 0.3979358315295815, "confidence_abs_error_epoch": 0.02344902089410559, "confidence_pred_mean_epoch": 0.09742121706091683, "confidence_cumprod_bias_epoch": 0.019025832961927846, "loss_epoch": 0.016075477810171688, "ce_loss_epoch": 0.025754998969284685, "tv_loss_epoch": 0.005977056000971818, "accept_rate_epoch": 0.6602440965940528, "accept_len_epoch": 2.2182613860095755, "full_acc_epoch": 0.6738053108479074, "position_1_acc_epoch": 0.7494481269932012, "position_2_acc_epoch": 0.5980975652161433}