michaelbenayoun
/

deepseekv3-tiny-4kv-heads-4-layers-random

Text Generation

text-generation-inference

Model card Files Files and versions

michaelbenayoun HF Staff commited on Jul 24

Commit

6f4bcbe

·

verified ·

1 Parent(s): 3592608

Upload DeepseekV3ForCausalLM

Files changed (2) hide show

config.json +1 -1
model.safetensors +2 -2

config.json CHANGED Viewed

@@ -35,7 +35,7 @@
   "num_nextn_predict_layers": 0,
   "pretraining_tp": 1,
   "q_lora_rank": 48,
-  "qk_head_dim": 48,
   "qk_nope_head_dim": 32,
   "qk_rope_head_dim": 32,
   "quantization_config": {

   "num_nextn_predict_layers": 0,
   "pretraining_tp": 1,
   "q_lora_rank": 48,
+  "qk_head_dim": 64,
   "qk_nope_head_dim": 32,
   "qk_rope_head_dim": 32,
   "quantization_config": {

model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1df2c380c0a617c83807445c3a2779128ba1677f2b8e2d9a794164cb4eacd1a2
-size 115530536

 version https://git-lfs.github.com/spec/v1
+oid sha256:bc77d5fa0f43af6e87ad6c1696b4463d169316adce893db57ae731bc268557ca
+size 115727160