v75 (S24/SM8650, soc 57): DeepSeek-R1-Distill-Qwen-1.5B fp16 bundle — export 8/8 (runtime needs >8GB RAM)
#2
by flying101 - opened
.gitattributes
CHANGED
|
@@ -35,3 +35,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
v81/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 37 |
v79/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
v81/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 37 |
v79/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
v75/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
v75/DeepSeek-R1-Distill-Qwen-1.5B.json
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"schema_version": 1,
|
| 3 |
+
"_comment": "",
|
| 4 |
+
"model": {
|
| 5 |
+
"name": "DeepSeek-R1-Distill-Qwen-1.5B",
|
| 6 |
+
"family": "llm",
|
| 7 |
+
"dsp_arch": "v75",
|
| 8 |
+
"tokenizer_pre": "qwen2"
|
| 9 |
+
},
|
| 10 |
+
"params": {
|
| 11 |
+
"hidden": 1536,
|
| 12 |
+
"vocab": 151936,
|
| 13 |
+
"n_layers": 28,
|
| 14 |
+
"max_ctx": 1024,
|
| 15 |
+
"kv_dim": 256,
|
| 16 |
+
"head_dim": 128,
|
| 17 |
+
"rope_theta": 10000.0,
|
| 18 |
+
"eos_token_id": 151643
|
| 19 |
+
},
|
| 20 |
+
"artifacts": {
|
| 21 |
+
"contexts": {
|
| 22 |
+
"prefill": {
|
| 23 |
+
"bin": "deepseekr1distillqwen15b_prefill_f16.bin"
|
| 24 |
+
},
|
| 25 |
+
"decode": {
|
| 26 |
+
"bin": "deepseekr1distillqwen15b_decode_f16.bin"
|
| 27 |
+
},
|
| 28 |
+
"lmhead": {
|
| 29 |
+
"bin": "deepseekr1distillqwen15b_lmhead_f16.bin"
|
| 30 |
+
}
|
| 31 |
+
},
|
| 32 |
+
"embed": "deepseekr1distillqwen15b_embed_f16.bin",
|
| 33 |
+
"tokenizer": "tokenizer.json"
|
| 34 |
+
},
|
| 35 |
+
"plan": {
|
| 36 |
+
"steps": [
|
| 37 |
+
{
|
| 38 |
+
"host": "qwen3_generate",
|
| 39 |
+
"params": {
|
| 40 |
+
"prefill": "deepseekr1distillqwen15b_prefill_f16",
|
| 41 |
+
"decode": "deepseekr1distillqwen15b_decode_f16",
|
| 42 |
+
"lmhead": "deepseekr1distillqwen15b_lmhead_f16",
|
| 43 |
+
"max_new": 64,
|
| 44 |
+
"neg": -50000
|
| 45 |
+
}
|
| 46 |
+
}
|
| 47 |
+
]
|
| 48 |
+
}
|
| 49 |
+
}
|
v75/deepseekr1distillqwen15b_decode_f16.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a28b61a2ff52ac990438401c19b26c78cb0ff45696fd49dfb7adf8ba56dc88a
|
| 3 |
+
size 2625544192
|
v75/deepseekr1distillqwen15b_embed_f16.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e55610c68685326d482c594ff3bb16141e71a0d219fe729211562ab630953c6e
|
| 3 |
+
size 466747392
|
v75/deepseekr1distillqwen15b_lmhead_f16.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:748d5f7623fa3292a2984533bb6392eddb7bedb26dec1d469d77c97e1881de56
|
| 3 |
+
size 467566592
|
v75/deepseekr1distillqwen15b_prefill_f16.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3e814ab96d841768c001995d238ea38d8829ec97feab16586167b013559df5c
|
| 3 |
+
size 2638245888
|
v75/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f624f8136bc57382156e7595e8ea404a66fbce12e3181b933ec2459bda3a71be
|
| 3 |
+
size 11421907
|