Add sample code
#4
by
Xenova
HF Staff
- opened
README.md
CHANGED
|
@@ -44,6 +44,355 @@ The Ministral 3 3B Instruct model offers the following capabilities:
|
|
| 44 |
- **Apache 2.0 License**: Open-source license allowing usage and modification for both commercial and non-commercial purposes.
|
| 45 |
- **Large Context Window**: Supports a 256k context window.
|
| 46 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
## License
|
| 48 |
|
| 49 |
This model is licensed under the [Apache 2.0 License](https://www.apache.org/licenses/LICENSE-2.0.txt).
|
|
|
|
| 44 |
- **Apache 2.0 License**: Open-source license allowing usage and modification for both commercial and non-commercial purposes.
|
| 45 |
- **Large Context Window**: Supports a 256k context window.
|
| 46 |
|
| 47 |
+
|
| 48 |
+
## Usage
|
| 49 |
+
|
| 50 |
+
|
| 51 |
+
### Transformers.js
|
| 52 |
+
|
| 53 |
+
#### Online WebGPU Demo
|
| 54 |
+
|
| 55 |
+
You can test out the model directly in your browser using our online WebGPU demo: https://huggingface.co/spaces/mistralai/Ministral_3B_WebGPU.
|
| 56 |
+
|
| 57 |
+
<video src="https://cdn-uploads.huggingface.co/production/uploads/61b253b7ac5ecaae3d1efe0c/LA0XUsH9pFZaphbfF-HGm.mp4" controls autoplay muted></video>
|
| 58 |
+
|
| 59 |
+
#### JavaScript
|
| 60 |
+
|
| 61 |
+
If you haven't already, you can install the [Transformers.js](https://huggingface.co/docs/transformers.js) JavaScript library from [NPM](https://www.npmjs.com/package/@huggingface/transformers) using:
|
| 62 |
+
```bash
|
| 63 |
+
npm i @huggingface/transformers
|
| 64 |
+
```
|
| 65 |
+
|
| 66 |
+
You can then run the model as follows:
|
| 67 |
+
```js
|
| 68 |
+
import {
|
| 69 |
+
AutoProcessor,
|
| 70 |
+
AutoModelForImageTextToText,
|
| 71 |
+
load_image,
|
| 72 |
+
TextStreamer,
|
| 73 |
+
} from "@huggingface/transformers";
|
| 74 |
+
|
| 75 |
+
// Load processor and model
|
| 76 |
+
const model_id = "mistralai/Ministral-3-3B-Instruct-2512-ONNX";
|
| 77 |
+
const processor = await AutoProcessor.from_pretrained(model_id);
|
| 78 |
+
const model = await AutoModelForImageTextToText.from_pretrained(model_id, {
|
| 79 |
+
device: "webgpu",
|
| 80 |
+
});
|
| 81 |
+
|
| 82 |
+
// Prepare inputs
|
| 83 |
+
const messages = [
|
| 84 |
+
{
|
| 85 |
+
role: "user",
|
| 86 |
+
content: [
|
| 87 |
+
{ type: "image" },
|
| 88 |
+
{
|
| 89 |
+
type: "text",
|
| 90 |
+
text: "What action do you think I should take in this situation? List all the possible actions and explain why you think they are good or bad.",
|
| 91 |
+
},
|
| 92 |
+
],
|
| 93 |
+
},
|
| 94 |
+
]
|
| 95 |
+
const prompt = processor.apply_chat_template(messages);
|
| 96 |
+
const url = "https://static.wikia.nocookie.net/essentialsdocs/images/7/70/Battle.png/revision/latest?cb=20220523172438";
|
| 97 |
+
const image = await load_image(url);
|
| 98 |
+
const inputs = await processor(image, prompt, { add_special_tokens: false });
|
| 99 |
+
|
| 100 |
+
// Generate response
|
| 101 |
+
const outputs = await model.generate({
|
| 102 |
+
...inputs,
|
| 103 |
+
max_new_tokens: 2048,
|
| 104 |
+
streamer: new TextStreamer(processor.tokenizer, {
|
| 105 |
+
skip_prompt: true,
|
| 106 |
+
// callback_function: (text) => { /* Do something with the streamed output */ },
|
| 107 |
+
}),
|
| 108 |
+
});
|
| 109 |
+
|
| 110 |
+
// Decode output
|
| 111 |
+
const decoded = processor.batch_decode(
|
| 112 |
+
outputs.slice(null, [inputs.input_ids.dims.at(-1), null]),
|
| 113 |
+
{ skip_special_tokens: true },
|
| 114 |
+
);
|
| 115 |
+
console.log(decoded[0]);
|
| 116 |
+
```
|
| 117 |
+
|
| 118 |
+
<details>
|
| 119 |
+
|
| 120 |
+
<summary>See example output</summary>
|
| 121 |
+
|
| 122 |
+
```
|
| 123 |
+
In this Pokémon game screenshot, you have several potential actions to consider, depending on the situation and your strategy. Here are the possible actions and their implications:
|
| 124 |
+
|
| 125 |
+
---
|
| 126 |
+
|
| 127 |
+
### **1. FIGHT**
|
| 128 |
+
**Description:** Use the Pikachu in your possession to attack the Pidgey.
|
| 129 |
+
|
| 130 |
+
#### **Pros:**
|
| 131 |
+
- **Potential to defeat Pidgey:** If Pikachu has strong moves (e.g., Thunderbolt, Electric Move) and Pidgey is weak to Electric-type attacks, this could be a strong choice.
|
| 132 |
+
- **Experience Gain:** Winning battles typically rewards you with experience points (XP) and sometimes drops Pidgey as a captured Pokémon.
|
| 133 |
+
- **Training Opportunity:** Helps you level up Pikachu, which can improve its stats and access new moves.
|
| 134 |
+
|
| 135 |
+
#### **Cons:**
|
| 136 |
+
- **Risk of Losing Pikachu:** If Pikachu is already at a low HP (e.g., 83/83 is not bad, but if it's lower), a loss could be detrimental.
|
| 137 |
+
- **Move Selection Matters:** If Pikachu doesn’t have a strong move against Pidgey, you might lose the battle unnecessarily.
|
| 138 |
+
- **Stamina Cost:** Battles can consume stamina if you're on a timer-based system (e.g., Pokémon Red/Blue).
|
| 139 |
+
|
| 140 |
+
---
|
| 141 |
+
|
| 142 |
+
### **2. RUN**
|
| 143 |
+
**Description:** Use the "Run" option to escape the battle.
|
| 144 |
+
|
| 145 |
+
#### **Pros:**
|
| 146 |
+
- **Avoids Risk:** If you're unsure about Pikachu’s moves or HP, running could save you from a potentially bad outcome.
|
| 147 |
+
- **Preserves Resources:** Running avoids losing Pikachu’s HP, which could be useful if you're running low.
|
| 148 |
+
- **Flexibility:** Allows you to explore or use other Pokémon later.
|
| 149 |
+
|
| 150 |
+
#### **Cons:**
|
| 151 |
+
- **No XP Gain:** Running means you won’t earn XP for defeating Pidgey.
|
| 152 |
+
- **Potential Consequences:** If you run too often, you might miss out on capturing Pidgey or gaining experience.
|
| 153 |
+
- **No Training:** Pikachu won’t level up or gain new moves if you avoid battles.
|
| 154 |
+
|
| 155 |
+
---
|
| 156 |
+
|
| 157 |
+
### **3. POKEMON (Select Another Pokémon)**
|
| 158 |
+
**Description:** If you have another Pokémon in your party, you could switch to it.
|
| 159 |
+
|
| 160 |
+
#### **Pros:**
|
| 161 |
+
- **Better Move Selection:** If you have a stronger Pokémon (e.g., a Fire-type or Ground-type) that can handle Pidgey, switching could be beneficial.
|
| 162 |
+
- **Balanced Strategy:** Helps you manage your team better.
|
| 163 |
+
- **Avoids Weaknesses:** If Pikachu is weak to something Pidgey might use, switching could prevent a loss.
|
| 164 |
+
|
| 165 |
+
#### **Cons:**
|
| 166 |
+
- **No Immediate XP Gain:** Switching doesn’t defeat Pidgey, so you won’t earn XP.
|
| 167 |
+
- **Stamina Cost:** Switching might still consume stamina if you're on a timer.
|
| 168 |
+
- **Potential to Lose Pikachu:** If you switch and Pikachu is already at a low HP, you might lose it.
|
| 169 |
+
|
| 170 |
+
---
|
| 171 |
+
|
| 172 |
+
### **Best Strategy Based on the Image:**
|
| 173 |
+
- **If you want to capture Pidgey:** Fighting is the best option, as it rewards you with experience and potentially captures Pidgey.
|
| 174 |
+
- **If you're unsure about Pikachu’s moves:** Running could be safer, but you might miss out on XP.
|
| 175 |
+
- **If you have another Pokémon:** Switching could be a good idea if you have a stronger option.
|
| 176 |
+
|
| 177 |
+
Would you like to know what moves Pikachu might have, or do you have another Pokémon in your party? That could help decide the best action!
|
| 178 |
+
```
|
| 179 |
+
|
| 180 |
+
</details>
|
| 181 |
+
|
| 182 |
+
|
| 183 |
+
### ONNXRuntime
|
| 184 |
+
|
| 185 |
+
```py
|
| 186 |
+
from transformers import AutoConfig, AutoProcessor
|
| 187 |
+
import onnxruntime
|
| 188 |
+
import numpy as np
|
| 189 |
+
from huggingface_hub import hf_hub_download
|
| 190 |
+
|
| 191 |
+
# 1. Load config, processor, and model
|
| 192 |
+
model_id = "mistralai/Ministral-3-3B-Instruct-2512-ONNX"
|
| 193 |
+
config = AutoConfig.from_pretrained(model_id)
|
| 194 |
+
processor = AutoProcessor.from_pretrained(model_id)
|
| 195 |
+
|
| 196 |
+
vision_model_path = hf_hub_download(model_id, "vision_encoder_q4.onnx", subfolder="onnx") # Download vision graph
|
| 197 |
+
hf_hub_download(model_id, "vision_encoder_q4.onnx_data", subfolder="onnx") # Download vision weights
|
| 198 |
+
embed_model_path = hf_hub_download(model_id, "embed_tokens_fp16.onnx", subfolder="onnx") # Download embed_tokens graph
|
| 199 |
+
hf_hub_download(model_id, "embed_tokens_fp16.onnx_data", subfolder="onnx") # Download embed_tokens weights
|
| 200 |
+
decoder_model_path = hf_hub_download(model_id, "decoder_model_merged_q4.onnx", subfolder="onnx") # Download decoder graph
|
| 201 |
+
hf_hub_download(model_id, "decoder_model_merged_q4.onnx_data", subfolder="onnx") # Download decoder weights (1/2)
|
| 202 |
+
hf_hub_download(model_id, "decoder_model_merged_q4.onnx_data_1", subfolder="onnx") # Download decoder weights (2/2)
|
| 203 |
+
|
| 204 |
+
## Load sessions
|
| 205 |
+
providers = ['CPUExecutionProvider']
|
| 206 |
+
vision_session = onnxruntime.InferenceSession(vision_model_path, providers=providers)
|
| 207 |
+
embed_session = onnxruntime.InferenceSession(embed_model_path, providers=providers)
|
| 208 |
+
decoder_session = onnxruntime.InferenceSession(decoder_model_path, providers=providers)
|
| 209 |
+
|
| 210 |
+
## Set config values
|
| 211 |
+
text_config = config.text_config
|
| 212 |
+
num_key_value_heads = text_config.num_key_value_heads
|
| 213 |
+
head_dim = text_config.head_dim
|
| 214 |
+
num_hidden_layers = text_config.num_hidden_layers
|
| 215 |
+
eos_token_id = text_config.eos_token_id
|
| 216 |
+
image_token_index = config.image_token_index
|
| 217 |
+
|
| 218 |
+
# 2. Prepare inputs
|
| 219 |
+
image_url = "https://static.wikia.nocookie.net/essentialsdocs/images/7/70/Battle.png/revision/latest?cb=20220523172438"
|
| 220 |
+
messages = [
|
| 221 |
+
{
|
| 222 |
+
"role": "user",
|
| 223 |
+
"content": [
|
| 224 |
+
{
|
| 225 |
+
"type": "text",
|
| 226 |
+
"text": "What action do you think I should take in this situation? List all the possible actions and explain why you think they are good or bad.",
|
| 227 |
+
},
|
| 228 |
+
{"type": "image", "url": image_url},
|
| 229 |
+
],
|
| 230 |
+
},
|
| 231 |
+
]
|
| 232 |
+
inputs = processor.apply_chat_template(messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt")
|
| 233 |
+
|
| 234 |
+
input_ids = inputs['input_ids'].numpy()
|
| 235 |
+
attention_mask = inputs['attention_mask'].numpy()
|
| 236 |
+
pixel_values = inputs['pixel_values'].numpy()
|
| 237 |
+
batch_size = input_ids.shape[0]
|
| 238 |
+
past_key_values = {
|
| 239 |
+
f'past_key_values.{layer}.{kv}': np.zeros([batch_size, num_key_value_heads, 0, head_dim], dtype=np.float32)
|
| 240 |
+
for layer in range(num_hidden_layers)
|
| 241 |
+
for kv in ('key', 'value')
|
| 242 |
+
}
|
| 243 |
+
position_ids = np.tile(np.arange(0, input_ids.shape[-1]), (batch_size, 1))
|
| 244 |
+
|
| 245 |
+
# 3. Generation loop
|
| 246 |
+
max_new_tokens = 1024
|
| 247 |
+
generated_tokens = np.array([[]], dtype=np.int64)
|
| 248 |
+
image_features = None
|
| 249 |
+
for i in range(max_new_tokens):
|
| 250 |
+
inputs_embeds = embed_session.run(None, {'input_ids': input_ids})[0]
|
| 251 |
+
|
| 252 |
+
if image_features is None:
|
| 253 |
+
## Only compute vision features if not already computed
|
| 254 |
+
image_features = vision_session.run(None, dict(
|
| 255 |
+
pixel_values=pixel_values,
|
| 256 |
+
))[0]
|
| 257 |
+
|
| 258 |
+
## Merge text and vision embeddings
|
| 259 |
+
inputs_embeds[input_ids == image_token_index] = image_features.reshape(-1, image_features.shape[-1])
|
| 260 |
+
|
| 261 |
+
logits, *present_key_values = decoder_session.run(None, dict(
|
| 262 |
+
inputs_embeds=inputs_embeds,
|
| 263 |
+
attention_mask=attention_mask,
|
| 264 |
+
position_ids=position_ids,
|
| 265 |
+
**past_key_values,
|
| 266 |
+
))
|
| 267 |
+
|
| 268 |
+
## Update values for next generation loop
|
| 269 |
+
input_ids = logits[:, -1].argmax(-1, keepdims=True)
|
| 270 |
+
attention_mask = np.concatenate([attention_mask, np.ones((batch_size, 1), dtype=attention_mask.dtype)], axis=-1)
|
| 271 |
+
position_ids = position_ids[:, -1:] + 1
|
| 272 |
+
for j, key in enumerate(past_key_values):
|
| 273 |
+
past_key_values[key] = present_key_values[j]
|
| 274 |
+
|
| 275 |
+
generated_tokens = np.concatenate([generated_tokens, input_ids], axis=-1)
|
| 276 |
+
if (input_ids == eos_token_id).all():
|
| 277 |
+
break
|
| 278 |
+
|
| 279 |
+
## (Optional) Streaming
|
| 280 |
+
print(processor.decode(input_ids[0]), end='', flush=True)
|
| 281 |
+
print()
|
| 282 |
+
|
| 283 |
+
# 4. Output result
|
| 284 |
+
print(processor.batch_decode(generated_tokens, skip_special_tokens=True)[0])
|
| 285 |
+
```
|
| 286 |
+
|
| 287 |
+
<details>
|
| 288 |
+
<summary>See example output</summary>
|
| 289 |
+
|
| 290 |
+
```
|
| 291 |
+
In this *Pokémon* game screenshot, you are presented with a battle scenario between **Pidgey** and **Pikachu**. Here are the possible actions along with an analysis of their potential outcomes:
|
| 292 |
+
|
| 293 |
+
---
|
| 294 |
+
|
| 295 |
+
### **1. FIGHT**
|
| 296 |
+
**Explanation:**
|
| 297 |
+
- You choose to engage in battle.
|
| 298 |
+
- Pidgey is a **Level 17** Pokémon with basic stats, while Pikachu is a **Level 42** Pokémon with significantly higher HP, Attack, and overall power.
|
| 299 |
+
- **Pros:**
|
| 300 |
+
- Pikachu is stronger and likely to win if it lands a hit.
|
| 301 |
+
- Pidgey may be able to land a few moves before being overwhelmed.
|
| 302 |
+
- **Cons:**
|
| 303 |
+
- Pidgey is likely to lose quickly due to its lower stats and experience.
|
| 304 |
+
- Pikachu may have a high chance of winning in one or two turns.
|
| 305 |
+
|
| 306 |
+
---
|
| 307 |
+
|
| 308 |
+
### **2. BAG**
|
| 309 |
+
**Explanation:**
|
| 310 |
+
- You choose to **Bag** Pikachu.
|
| 311 |
+
- **Pros:**
|
| 312 |
+
- You can keep Pikachu in your bag for later use.
|
| 313 |
+
- If you need to switch Pokémon or use Pikachu in another battle, this is a viable option.
|
| 314 |
+
- **Cons:**
|
| 315 |
+
- You miss out on the opportunity to potentially defeat Pidgey and gain experience.
|
| 316 |
+
- If Pidgey is a rare or special Pokémon, bagging it might not be ideal.
|
| 317 |
+
|
| 318 |
+
---
|
| 319 |
+
|
| 320 |
+
### **3. RUN**
|
| 321 |
+
**Explanation:**
|
| 322 |
+
- You choose to **Run** (escape the battle).
|
| 323 |
+
- **Pros:**
|
| 324 |
+
- You avoid losing HP and experience points.
|
| 325 |
+
- You can continue your journey without facing a weaker opponent.
|
| 326 |
+
- **Cons:**
|
| 327 |
+
- You might miss out on a chance to level up Pikachu or gain experience.
|
| 328 |
+
- If you are in a competitive setting, running might not be the best strategy.
|
| 329 |
+
|
| 330 |
+
---
|
| 331 |
+
|
| 332 |
+
### **Summary of Best Options:**
|
| 333 |
+
- **If you want to continue the game and gain experience:**
|
| 334 |
+
- **Fight** Pikachu, but be prepared for a potentially quick loss.
|
| 335 |
+
- **If you want to save Pikachu for later:**
|
| 336 |
+
- **Bag** Pikachu and continue your journey.
|
| 337 |
+
- **If you want to avoid unnecessary battles:**
|
| 338 |
+
- **Run** away from the battle.
|
| 339 |
+
|
| 340 |
+
Since Pikachu is significantly stronger, **Fighting** is the most straightforward choice, but you might want to consider **Bagging** Pikachu if you want to keep it for later battles. If you're not in a rush, **Running** is also a viable option.
|
| 341 |
+
In this *Pokémon* game screenshot, you are presented with a battle scenario between **Pidgey** and **Pikachu**. Here are the possible actions along with an analysis of their potential outcomes:
|
| 342 |
+
|
| 343 |
+
---
|
| 344 |
+
|
| 345 |
+
### **1. FIGHT**
|
| 346 |
+
**Explanation:**
|
| 347 |
+
- You choose to engage in battle.
|
| 348 |
+
- Pidgey is a **Level 17** Pokémon with basic stats, while Pikachu is a **Level 42** Pokémon with significantly higher HP, Attack, and overall power.
|
| 349 |
+
- **Pros:**
|
| 350 |
+
- Pikachu is stronger and likely to win if it lands a hit.
|
| 351 |
+
- Pidgey may be able to land a few moves before being overwhelmed.
|
| 352 |
+
- **Cons:**
|
| 353 |
+
- Pidgey is likely to lose quickly due to its lower stats and experience.
|
| 354 |
+
- Pikachu may have a high chance of winning in one or two turns.
|
| 355 |
+
|
| 356 |
+
---
|
| 357 |
+
|
| 358 |
+
### **2. BAG**
|
| 359 |
+
**Explanation:**
|
| 360 |
+
- You choose to **Bag** Pikachu.
|
| 361 |
+
- **Pros:**
|
| 362 |
+
- You can keep Pikachu in your bag for later use.
|
| 363 |
+
- If you need to switch Pokémon or use Pikachu in another battle, this is a viable option.
|
| 364 |
+
- **Cons:**
|
| 365 |
+
- You miss out on the opportunity to potentially defeat Pidgey and gain experience.
|
| 366 |
+
- If Pidgey is a rare or special Pokémon, bagging it might not be ideal.
|
| 367 |
+
|
| 368 |
+
---
|
| 369 |
+
|
| 370 |
+
### **3. RUN**
|
| 371 |
+
**Explanation:**
|
| 372 |
+
- You choose to **Run** (escape the battle).
|
| 373 |
+
- **Pros:**
|
| 374 |
+
- You avoid losing HP and experience points.
|
| 375 |
+
- You can continue your journey without facing a weaker opponent.
|
| 376 |
+
- **Cons:**
|
| 377 |
+
- You might miss out on a chance to level up Pikachu or gain experience.
|
| 378 |
+
- If you are in a competitive setting, running might not be the best strategy.
|
| 379 |
+
|
| 380 |
+
---
|
| 381 |
+
|
| 382 |
+
### **Summary of Best Options:**
|
| 383 |
+
- **If you want to continue the game and gain experience:**
|
| 384 |
+
- **Fight** Pikachu, but be prepared for a potentially quick loss.
|
| 385 |
+
- **If you want to save Pikachu for later:**
|
| 386 |
+
- **Bag** Pikachu and continue your journey.
|
| 387 |
+
- **If you want to avoid unnecessary battles:**
|
| 388 |
+
- **Run** away from the battle.
|
| 389 |
+
|
| 390 |
+
Since Pikachu is significantly stronger, **Fighting** is the most straightforward choice, but you might want to consider **Bagging** Pikachu if you want to keep it for later battles. If you're not in a rush, **Running** is also a viable option.
|
| 391 |
+
```
|
| 392 |
+
|
| 393 |
+
</details>
|
| 394 |
+
|
| 395 |
+
|
| 396 |
## License
|
| 397 |
|
| 398 |
This model is licensed under the [Apache 2.0 License](https://www.apache.org/licenses/LICENSE-2.0.txt).
|