How use this model ?
hi
how use this ?
Z:\llama-b3778-bin-win-avx512-x64>llama-cli --hf-repo tspersian/mT5_base_translation_English_to_Persian-Farsi-Q4_K_M-GGUF --hf-file mt5_base_translation_english_to_persian-farsi-q4_k_m.gguf -p "The meaning to life and the universe is"
Z:\llama-b3778-bin-win-avx512-x64>llama-cli --hf-repo tspersian/mT5_base_translation_English_to_Persian-Farsi-Q4_K_M-GGUF --hf-file path/to/mt5_base_translation_english_to_persian-farsi-q4_k_m.gguf -p "Test prompt"
Z:\llama-b3778-bin-win-avx512-x64>llama-cli --hf-repo tspersian/mT5_base_translation_English_to_Persian-Farsi-Q4_K_M-GGUF --hf-file mt5_base_translation_english_to_persian-farsi-q4_k_m.gguf -p "The meaning to life and the universe is"
using LLama;
using LLama.Common;
using System.Threading.Tasks;
string modelPath = @"z://mt5_base_translation_english_to_persian-farsi-q4_k_m.gguf"; // Model path for the downloaded GGUF model
var parameters = new ModelParams(modelPath)
{
ContextSize = 1024,
GpuLayerCount = 5
};
using var model = LLamaWeights.LoadFromFile(parameters);
using var context = model.CreateContext(parameters);
var executor = new InstructExecutor(context);
string prompt = "Translate this sentence to Persian.";
await foreach (var token in executor.InferAsync(prompt, new InferenceParams { MaxTokens = 256 }))
{
Console.Write(token); // Stream output token by token
}
Console.WriteLine();
llama_model_loader: loaded meta data with 27 key-value pairs and 282 tensors from z://mt5_base_translation_english_to_persian-farsi-q4_k_m.gguf (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = t5
llama_model_loader: - kv 1: general.name str = T5
llama_model_loader: - kv 2: t5.context_length u32 = 512
llama_model_loader: - kv 3: t5.embedding_length u32 = 768
llama_model_loader: - kv 4: t5.feed_forward_length u32 = 2048
llama_model_loader: - kv 5: t5.block_count u32 = 12
llama_model_loader: - kv 6: t5.attention.head_count u32 = 12
llama_model_loader: - kv 7: t5.attention.key_length u32 = 64
llama_model_loader: - kv 8: t5.attention.value_length u32 = 64
llama_model_loader: - kv 9: t5.attention.layer_norm_epsilon f32 = 0.000001
llama_model_loader: - kv 10: t5.attention.relative_buckets_count u32 = 32
llama_model_loader: - kv 11: t5.attention.layer_norm_rms_epsilon f32 = 0.000001
llama_model_loader: - kv 12: t5.decoder_start_token_id u32 = 0
llama_model_loader: - kv 13: general.file_type u32 = 15
llama_model_loader: - kv 14: tokenizer.ggml.model str = t5
llama_model_loader: - kv 15: tokenizer.ggml.pre str = default
llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,250112] = ["", "", "", "<0x00>", ...
llama_model_loader: - kv 17: tokenizer.ggml.scores arr[f32,250112] = [0.000000, 0.000000, 0.000000, 0.0000...
llama_model_loader: - kv 18: tokenizer.ggml.token_type arr[i32,250112] = [3, 3, 2, 6, 6, 6, 6, 6, 6, 6, 6, 6, ...
llama_model_loader: - kv 19: tokenizer.ggml.add_space_prefix bool = true
llama_model_loader: - kv 20: tokenizer.ggml.remove_extra_whitespaces bool = true
llama_model_loader: - kv 21: tokenizer.ggml.precompiled_charsmap arr[u8,237539] = [0, 180, 2, 0, 0, 132, 0, 0, 0, 0, 0,...
llama_model_loader: - kv 22: tokenizer.ggml.eos_token_id u32 = 1
llama_model_loader: - kv 23: tokenizer.ggml.padding_token_id u32 = 0
llama_model_loader: - kv 24: tokenizer.ggml.add_bos_token bool = false
llama_model_loader: - kv 25: tokenizer.ggml.add_eos_token bool = true
llama_model_loader: - kv 26: general.quantization_version u32 = 2
llama_model_loader: - type f32: 62 tensors
llama_model_loader: - type f16: 2 tensors
llama_model_loader: - type q4_K: 181 tensors
llama_model_loader: - type q6_K: 37 tensors
llm_load_vocab: special tokens cache size = 15
llm_load_vocab: token to piece cache size = 1.6229 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = t5
llm_load_print_meta: vocab type = UGM
llm_load_print_meta: n_vocab = 250112
llm_load_print_meta: n_merges = 0
llm_load_print_meta: vocab_only = 0
llm_load_print_meta: n_ctx_train = 512
llm_load_print_meta: n_embd = 768
llm_load_print_meta: n_layer = 12
llm_load_print_meta: n_head = 12
llm_load_print_meta: n_head_kv = 12
llm_load_print_meta: n_rot = 64
llm_load_print_meta: n_swa = 0
llm_load_print_meta: n_embd_head_k = 64
llm_load_print_meta: n_embd_head_v = 64
llm_load_print_meta: n_gqa = 1
llm_load_print_meta: n_embd_k_gqa = 768
llm_load_print_meta: n_embd_v_gqa = 768
llm_load_print_meta: f_norm_eps = 0.0e+00
llm_load_print_meta: f_norm_rms_eps = 1.0e-06
llm_load_print_meta: f_clamp_kqv = 0.0e+00
llm_load_print_meta: f_max_alibi_bias = 0.0e+00
llm_load_print_meta: f_logit_scale = 0.0e+00
llm_load_print_meta: n_ff = 2048
llm_load_print_meta: n_expert = 0
llm_load_print_meta: n_expert_used = 0
llm_load_print_meta: causal attn = 1
llm_load_print_meta: pooling type = 0
llm_load_print_meta: rope type = -1
llm_load_print_meta: rope scaling = linear
llm_load_print_meta: freq_base_train = 10000.0
llm_load_print_meta: freq_scale_train = 1
llm_load_print_meta: n_ctx_orig_yarn = 512
llm_load_print_meta: rope_finetuned = unknown
llm_load_print_meta: ssm_d_conv = 0
llm_load_print_meta: ssm_d_inner = 0
llm_load_print_meta: ssm_d_state = 0
llm_load_print_meta: ssm_dt_rank = 0
llm_load_print_meta: ssm_dt_b_c_rms = 0
llm_load_print_meta: model type = 250M
llm_load_print_meta: model ftype = Q4_K - Medium
llm_load_print_meta: model params = 582.40 M
llm_load_print_meta: model size = 369.38 MiB (5.32 BPW)
llm_load_print_meta: general.name = T5
llm_load_print_meta: EOS token = 1 ''
llm_load_print_meta: UNK token = 2 ''
llm_load_print_meta: PAD token = 0 ''
llm_load_print_meta: LF token = 259 '螕没眉'
llm_load_print_meta: max token length = 48
llm_load_tensors: ggml ctx size = 0.11 MiB
llm_load_tensors: offloading 5 repeating layers to GPU
llm_load_tensors: offloaded 5/13 layers to GPU
llm_load_tensors: CPU buffer size = 369.38 MiB
..................................
llama_new_context_with_model: n_ctx = 1024
llama_new_context_with_model: n_batch = 512
llama_new_context_with_model: n_ubatch = 512
llama_new_context_with_model: flash_attn = 0
llama_new_context_with_model: freq_base = 10000.0
llama_new_context_with_model: freq_scale = 1
llama_kv_cache_init: CPU KV buffer size = 36.00 MiB
llama_new_context_with_model: KV self size = 36.00 MiB, K (f16): 18.00 MiB, V (f16): 18.00 MiB
llama_new_context_with_model: CPU output buffer size = 0.95 MiB
llama_new_context_with_model: CPU compute buffer size = 45.50 MiB
llama_new_context_with_model: graph nodes = 425
llama_new_context_with_model: graph splits = 1
D:\a\LLamaSharp\LLamaSharp\src\llama.cpp:14039: GGML_ASSERT(n_outputs_enc > 0 && "call llama_encode() first") failed
z:\ConsoleApp1\ConsoleApp1\bin\Debug\net8.0\ConsoleApp1.exe (process 5636) exited with code -1073740791.
Press any key to close this window . . .
hello Alireza
here is brief summery how you can run it in python:
Install necessary packages:
pip install transformers torch
Python code:
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
# Load the tokenizer and the model from Hugging Face repository
model_name = "tspersian/mT5_base_translation_English_to_Persian-Farsi"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
# Function to translate English to Persian
def translate(text):
inputs = tokenizer(text, return_tensors="pt", padding=True, truncation=True)
output_ids = model.generate(inputs['input_ids'], max_length=256)
translation = tokenizer.decode(output_ids[0], skip_special_tokens=True)
return translation
# Example prompt
prompt = "The meaning of life and the universe is"
translation = translate(prompt)
print(translation)
and if you want to use llama cli you can do this:
llama-cli --hf-repo tspersian/mT5_base_translation_English_to_Persian-Farsi-Q4_K_M-GGUF \
--hf-file /path/to/model/mt5_base_translation_english_to_persian-farsi-q4_k_m.gguf \
-p "The meaning of life and the universe is"
make sure "/path/to/model" is correct and model is placed there.