Hugging Face's logo

bunyaminergen
/
Qwen2.5-Coder-1.5B-Instruct-Reasoning

Text Generation
PEFT
Safetensors
English
code
codeqwen
chat
qwen
qwen-coder
conversational
Model card Files
xet
Community
bunyaminergen commited on
Commit
c092dc8
·
1 Parent(s): 939ee59

Initial

Browse files
Files changed (1) hide show
  1. README.md +25 -27
README.md CHANGED
@@ -137,44 +137,42 @@ print(tokenizer.decode(outputs[0], skip_special_tokens=True))
137
  | bnb_4bit_quant_type | `nf4` |
138
  | bnb_4bit_use_double_quant | `true` |
139
 
140
- Aşağıda her başlık için ayrı birer tablo oluşturdum:
141
-
142
  #### Dataset
143
 
144
- | Parameter | Value |
145
- |----------------------|--------------------------------|
146
- | Dataset Name | `nvidia/OpenCodeReasoning` |
147
- | Split | `split_0` |
148
- | Number of Rows | `8000` |
149
- | Max Token Length | `8192` |
150
- | Shuffle | `True` |
151
- | Number of Processes | `4` |
152
 
153
  #### Tokenizer
154
 
155
- | Parameter | Value |
156
- |--------------------------------|-------------------------------|
157
- | Truncation | Enabled (`max_length=8192`) |
158
- | Masked Language Modeling (MLM) | `False` |
159
 
160
  #### Speeds, Sizes, Times
161
 
162
- | Parameter | Value |
163
- |-------------------------|-------------------------------------------------------------|
164
- | Total Training Time | ~3.5 hours |
165
- | Checkpoint Frequency | every `10000` steps |
166
- | Checkpoint Steps | `checkpoint-10000`, `checkpoint-20000`, `checkpoint-24000` |
167
 
168
  #### Compute Infrastructure
169
 
170
- | Parameter | Value |
171
- |--------------|--------------------------------------|
172
- | GPU | 1 × NVIDIA H100 SXM (80 GB VRAM) |
173
- | RAM | 125 GB |
174
- | CPU | 16 vCPU |
175
- | OS | Ubuntu 22.04 |
176
- | Frameworks | PyTorch 2.4.0 |
177
- | CUDA Version | 12.4.1 |
178
 
179
  ---
180
 
 
137
  | bnb_4bit_quant_type | `nf4` |
138
  | bnb_4bit_use_double_quant | `true` |
139
 
 
 
140
  #### Dataset
141
 
142
+ | Parameter | Value |
143
+ |---------------------|----------------------------|
144
+ | Dataset Name | `nvidia/OpenCodeReasoning` |
145
+ | Split | `split_0` |
146
+ | Number of Rows | `8000` |
147
+ | Max Token Length | `8192` |
148
+ | Shuffle | `True` |
149
+ | Number of Processes | `4` |
150
 
151
  #### Tokenizer
152
 
153
+ | Parameter | Value |
154
+ |--------------------------------|-----------------------------|
155
+ | Truncation | Enabled (`max_length=8192`) |
156
+ | Masked Language Modeling (MLM) | `False` |
157
 
158
  #### Speeds, Sizes, Times
159
 
160
+ | Parameter | Value |
161
+ |----------------------|------------------------------------------------------------|
162
+ | Total Training Time | ~3.5 hours |
163
+ | Checkpoint Frequency | every `10000` steps |
164
+ | Checkpoint Steps | `checkpoint-10000`, `checkpoint-20000`, `checkpoint-24000` |
165
 
166
  #### Compute Infrastructure
167
 
168
+ | Parameter | Value |
169
+ |--------------|----------------------------------|
170
+ | GPU | 1 × NVIDIA H100 SXM (80 GB VRAM) |
171
+ | RAM | 125 GB |
172
+ | CPU | 16 vCPU |
173
+ | OS | Ubuntu 22.04 |
174
+ | Frameworks | PyTorch 2.4.0 |
175
+ | CUDA Version | 12.4.1 |
176
 
177
  ---
178