sentence-transformers How to use dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF with sentence-transformers:
from sentence_transformers import SentenceTransformer
model = SentenceTransformer("dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF")
sentences = [
"What is the most successful treatment for lung SFTs?\n",
"Recent progress in radioimmunoassay (RIA) technology has shown that thyroglobulin (Tg) is a protein normally secreted by the thyroid gland (Van Herle et al., 1979) and can be detected in the serum of virtually all normal subjects (Roitt and Torrigiani, 1967; Van Herle et al., 1973; Pezzino et al., 1977; Bodlaender et al., 1978; Roti et al., 1981) . The clinical usefulness of measuring serum Tg levels in thyroidal and nonthyroidal disorders has been pointed out by many workers. For example, high levels of serum Tg have been reported in patients with Graves' disease (Van Herle et al., 1973; Izumi and Larsen, 1978; Gardner et al., 1979; Pacini et al., 1980) , subacute thyroiditis (Van Herle et al., 1973; Izumi and Larsen, 1978; Gardner et al., 1979; Pacini et al., 1980) , and differentiated thyroid carcinoma (Van Herle and Uller, 1975; Shlossberg et al., 1979; Pacini et al., 1980) , whereas the suppression of serum Tg levels in thyrotoxicosis factitia has been reported (Mariotti et al., 1982) .\n\n Elevated levels of serum Tg have also been reported in normal pregnant women at the third trimester (Torrigiani et al., 1969) and at delivery (Van Herle et al., 1973; Pacini et al., 1980; Roti et al., 1981) . However, there has been little information concerning the level of serum Tg in each trimester of normal pregnancy. Therefore, it is not clear whether the levels of serum Tg in pregnant women can be used in the assessment of the clinical states of thyroidal disorders. subjects showed an approximately normal distribution in the logarithmic scale (Van Herle et al., 1976; Nakamura et al., 1984) . The Mann and Whitney U test was per formed using the native values for Tg (Pezzino et. al., 1977; Nakamura et al., 1984) . Differences were considered to be significant if p was less than 0.05. Mean values in the results were calculated from the levels of log Tg.\n\n Results Figure 1 shows the concentrations of serum Tg obtained from 52 pregnant and 15 nonpregnant females. Table 1 shows the mean logarithms for serum Tg and the range. The mean serum Tg concentration calculated from the logarithms for serum Tg of nonpregnant control females was 6.0ng/ml with a range of 1.5 to 23.6ng/ml.\n\n For pregnant women, the mean serum Tg levels at the first, second, early third, and the late third trimesters were 8.4 (range 1.3-18.0), 9.2 (1.7-25.6), 10.1 (1.8-22.8), and 12.1ng/ml (5.3-25.2), respectively. The variation of log Tg levels in the control group was not significantly different from that of the 4 pregnant groups.\n\n The mean value at the late third trimester was significantly higher (p<0.05 in both statistical analyses) than that of the controls. On the other hand, each mean value at the first, second, and early third \n\n Virtually all aspects of thyroid hormone economy are affected by pregnancy. For example, it is well known that T3 and T4 steadily increase during the first trimester and remain at high levels thereafter (Chan et al., 1975; Yamamoto et al., 1979; Skjoldebrand et al., 1982) .\n\n Concerning the values for serum Tg in normal pregnant women, only a few reports have been published.\n\n An earlier report by Hjort and Pedersen (1962) using a semiquantitative hemagglutination-inhibition technique showed that serum Tg was not detected in nonpregnant women but was detected in 3, 11, and 50% of pregnant women at the end of the first trimester, at six weeks before delivery, and at delivery, respectively. Torrigiani et al. (1969) , using RIA, compared the serum Tg levels of nonpregnant women with those of pregnant women at the first and third trimesters.",
"Les patients et familles qui ne sont pas en faveur de l'aide à mourir et ne comprennent pas en quoi cette approche diffère de l'abstention et l'interruption des traitements de maintien de la vie pourraient être réfractaires à l'idée d'interrompre un traitement de maintien de la vie lorsque cela est médicalement adapté. Ils pourraient avoir peur ou accuser les professionnels de la santé et les hôpitaux d'avoir pour but premier d'accélérer la mort, ne comprenant pas que les recommandations visant à limiter ou interrompre certaines interventions se fondent sur l'absence de bienfaits du traitement. D'autres paramètres contextuels, tels que la demande pour les services de l'USI et le nombre de lits, pourraient accentuer le manque de confiance. Alors que l'accélération intentionnelle du processus de mourir est illégale sauf dans le contexte très spécifique de l'aide à mourir, dans certains pays, des médecins 6 et infirmières ont rapporté cette pratique. 7 La Société belge de soins intensifs et de médecine d'urgence (SIZ) a récemment publié une déclaration soutenant l'utilisation de médicaments en l'absence d'inconfort du patient afin « d'améliorer la qualité de la mort », raccourcissant ainsi le processus de décès.\n\n 8 L'exposé de principe de la SCSI servira de fondement pour lancer la conversation concernant les pratiques actuelles d'abstention et d'interruption des traitements de maintien de la vie dans les USI canadiennes et la façon dont les recommandations pourront être appliquées dans la pratique afin de faciliter des soins de fin de vie transparents respectant les valeurs et préférences des patients tout en tenant compte des expériences des personnes prenant soin d'eux.\n\n Alors que les progrès en médecine et en chirurgie permettent aux patients atteints de maladie grave de vivre plus longtemps que jamais, les conversations portant sur les soins de fin de vie et les décisions concernant l'abstention et l'interruption des traitements de maintien de la vie vont devenir de plus en plus fréquentes. Dans le contexte d'un cadre clinique en évolution, les questions éthiques entourant l'abstention et l'interruption des traitements de maintien de la vie devront constamment être réévaluées. D'un point de vue très personnel, la mort d'un proche fait partie de l'intimité d'une relation. En tant que professionnels de la santé, nous avons le privilège d'être invités dans cette relation. Ce privilège s'accompagne d'une lourde responsabilité. L'intégrité, la compassion, et une approche interprofessionnelle fondée sur les principes face aux défis que comportent les conversations concernant l'abstention et l'interruption des traitements de maintien de la vie permettront que les soins, qu'il s'agisse de ceux jusqu'à ou entourant la fin de vie, soient personnalisés selon les besoins dynamiques et contextuels de nos patients et de leurs familles.\n\n Editorial responsibility This submission was handled by Dr. Hilary P. Grocott, Editor-in-Chief, Canadian Journal of Anesthesia.\n\n Financial support None.\n\n Responsabilité éditoriale Cet article a été traité par Dr Hilary P. Grocott, rédacteur en chef, Journal canadien d'anesthésie.\n\n Soutien financier Aucun.",
"2 The differential diagnosis of a lung SFT includes numerous malignant and benign tumors, including hemangiopericytoma, mesothelioma, monophasic synovial sarcoma, sarcomatoid renal cell carcinoma, spindle cell lipoma, fibrosarcoma, leiomyosarcoma and neurogenic tumors, including schwan-noma and malignant peripheral nerve sheath tumors. 6 Immunohistochemistry is significant for differentiating between these tumors. For example, leiomyosarcomas are CD34-negative and positive for SMA and desmin. Thus, the presence of different markers in leiomyosarcoma distinguishes it from an SFT.\n\n The most successful treatment for SFTs is surgery. In the present case, two surgical approaches were prepared depending on the intraoperative frozen section. If the tumor was benign, a total tumor excision was to be performed. A lower left lung lobe resection was to be performed for a malignant tumor, as well as a lymph node dissection involving radical dissection of the mediastinum. In the present case, the tumor was identified as benign. The efficacy of surgery depends on the entirety of the tumor resection. 8 Previous studies have reported the use of adjuvant radiotherapy or chemotherapy for malignant SFTs; however, the effectiveness of such treatment has yet to be elucidated. 3 Metastases are usually blood-borne and have been identified in the liver, bone, brain, lungs and muscles. 2 Chen et al. 9 reported the case of a 78-year-old male who presented with a left gluteal soft tissue mass with a twomonth history of a newly-diagnosed metastatic lung adenocarcinoma. H&E staining of the mass revealed an SFT containing metastases from the adenocarcinoma. 9 Furthermore, Schirosi et al. 10 reported that high p53 expression in SFTs may be significantly correlated with a poor prognosis. Due to the potential for metastasis and recurrence, long-term follow-up for several years is required. Watanabe et al. 11 described the case of a 57-year old female who was diagnosed with adenocarcinoma of the right, middle lung lobe and malignant visceral pleura SFT of the left, upper lung lobe. The patient received a median sternotomy. The study illustrated the importance of the diagnosis and treatment of two coexisting primary malignancies.\n\n In conclusion, the present study described a rare case of an SFT arising from the lung in a male patient. A tumor resection was successfully performed and the diagnosis of an SFT was determined based on the immunohistochemical findings. Long-term follow-up with radiological imaging is required to monitor the recurrence and metastasis of this type of tumor."
]
embeddings = model.encode(sentences)
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [4, 4] llama.cpp How to use dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF with llama.cpp:
Install (macOS, Linux) curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF:Q6_K
# Run inference directly in the terminal:
llama cli -hf dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF:Q6_K Install from WinGet (Windows) winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF:Q6_K
# Run inference directly in the terminal:
llama cli -hf dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF:Q6_K Use pre-built binary # Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF:Q6_K
# Run inference directly in the terminal:
./llama-cli -hf dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF:Q6_K Build from source code git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF:Q6_K
# Run inference directly in the terminal:
./build/bin/llama-cli -hf dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF:Q6_K Use Docker docker model run hf.co/dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF:Q6_K 
# !pip install llama-cpp-python from llama_cpp import Llama llm = Llama.from_pretrained( repo_id="dark-pen/EmbeddingsG300M-ft-Q6_K-GGUF", filename="embeddingsg300m-ft-q6_k.gguf", )