SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Fix : model used to test ggml conversion of Falcon-7b is incorrect (#35083)

Browse Source 
fixing test model
This commit is contained in:
Mohamed Mekkouri
2024-12-16 13:21:44 +01:00
committed by GitHub
parent 14910281a7
commit 85eb339231
1 changed files with 5 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
tests/quantization/ggml/test_ggml.py
View File

@@ -45,7 +45,8 @@ class GgufIntegrationTests(unittest.TestCase):
phi3_model_id = "microsoft/Phi-3-mini-4k-instruct-gguf" phi3_model_id = "microsoft/Phi-3-mini-4k-instruct-gguf"
bloom_model_id = "afrideva/bloom-560m-GGUF" bloom_model_id = "afrideva/bloom-560m-GGUF"
original_bloom_model_id = "bigscience/bloom-560m" original_bloom_model_id = "bigscience/bloom-560m"
falcon7b_model_id = "xaviviro/falcon-7b-quantized-gguf" falcon7b_model_id_q2 = "xaviviro/falcon-7b-quantized-gguf"
falcon7b_model_id_fp16 = "medmekk/falcon-7b-gguf"
falcon40b_model_id = "maddes8cht/tiiuae-falcon-40b-gguf" falcon40b_model_id = "maddes8cht/tiiuae-falcon-40b-gguf"
original_flacon7b_model_id = "tiiuae/falcon-7b" original_flacon7b_model_id = "tiiuae/falcon-7b"
t5_model_id = "repetitio/flan-t5-small" t5_model_id = "repetitio/flan-t5-small"
@@ -615,9 +616,9 @@ class GgufIntegrationTests(unittest.TestCase):
self.assertEqual(tokenizer.decode(out[0], skip_special_tokens=True), EXPECTED_TEXT) self.assertEqual(tokenizer.decode(out[0], skip_special_tokens=True), EXPECTED_TEXT)
def test_falcon7b_q2_k(self): def test_falcon7b_q2_k(self):
tokenizer = AutoTokenizer.from_pretrained(self.falcon7b_model_id, gguf_file=self.q2_k_falcon7b_model_id) tokenizer = AutoTokenizer.from_pretrained(self.falcon7b_model_id_q2, gguf_file=self.q2_k_falcon7b_model_id)
model = AutoModelForCausalLM.from_pretrained( model = AutoModelForCausalLM.from_pretrained(
self.falcon7b_model_id, self.falcon7b_model_id_q2,
gguf_file=self.q2_k_falcon7b_model_id, gguf_file=self.q2_k_falcon7b_model_id,
device_map="auto", device_map="auto",
torch_dtype=torch.float16, torch_dtype=torch.float16,
@@ -631,7 +632,7 @@ class GgufIntegrationTests(unittest.TestCase):
def test_falcon7b_weights_conversion_fp16(self): def test_falcon7b_weights_conversion_fp16(self):
quantized_model = AutoModelForCausalLM.from_pretrained( quantized_model = AutoModelForCausalLM.from_pretrained(
self.falcon7b_model_id, self.falcon7b_model_id_fp16,
gguf_file=self.fp16_falcon7b_model_id, gguf_file=self.fp16_falcon7b_model_id,
device_map="auto", device_map="auto",
torch_dtype=torch.float16, torch_dtype=torch.float16,